Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Table of Contents
Foreword 0
15
Part I Fathom 2 Help
1 Credits and ...................................................................................................................................
Acknowledgements 15
18
Part II What's New in Fathom 2?
23
Part III Getting Started
1 Getting Data................................................................................................................................... 24
Open a Fathom ..........................................................................................................................................................
Sample Document 24
Enter Data from
..........................................................................................................................................................
Scratch 25
Copy and Paste..........................................................................................................................................................
Data from Another Program 26
Have Fathom ..........................................................................................................................................................
Randomly Generate Data 27
Import Census ..........................................................................................................................................................
Microdata 30
2 Beginner Graphing
................................................................................................................................... 31
3 Undo and Redo
................................................................................................................................... 32
33
Part IV How To’s
1 Work with Graphs
................................................................................................................................... 33
Replace an Attribute
..........................................................................................................................................................
on a Graph with Another Attribute 35
Add Another Attribute
..........................................................................................................................................................
to an Already Occupied Axis 35
Add an Attribute
..........................................................................................................................................................
to the Middle of a Graph 36
Remove Attributes
..........................................................................................................................................................
from a Graph 37
Select Data in..........................................................................................................................................................
a Graph 37
Drag Data in a..........................................................................................................................................................
Graph 39
Change Axis Scales
.......................................................................................................................................................... 40
Change Axes.........................................................................................................................................................
by Dragging 40
Change Axes.........................................................................................................................................................
Numerically 41
Zoom In or .........................................................................................................................................................
Out in a Graph 41
Rearrange.........................................................................................................................................................
Categories on an Axis 42
Reverse a.........................................................................................................................................................
Graph’s Axis Scale 42
Dynamically Link
..........................................................................................................................................................
Axes to Each Other 43
Prevent Graph ..........................................................................................................................................................
Axes from Automatically Rescaling 45
Change Bin Width
.......................................................................................................................................................... 46
Change Bin .........................................................................................................................................................
Width by Dragging 46
Change Bin .........................................................................................................................................................
Width Numerically 47
Add Lines and..........................................................................................................................................................
Functions to Graphs 48
Add a Movable
.........................................................................................................................................................
Line to a Graph 48
Plot a Line.........................................................................................................................................................
at a Given Value on a Graph 49
Show Measures
.........................................................................................................................................................
of Spread in a Graph 51
Plot a Distribution
.........................................................................................................................................................
Over a Histogram 52
Add Fitted.........................................................................................................................................................
Lines to a Graph 53
Plot a Function
.........................................................................................................................................................
on a Graph 53
Plot Functions
.........................................................................................................................................................
on a Function Plot 54
Plot a Function
.........................................................................................................................................................
You Can Vary Dynamically (Using a Slider) 55
Trace on .........................................................................................................................................................
a Function or Fitted Line 56
Edit a Plotted
.........................................................................................................................................................
Function 56
2
Contents
Copy Functions
.........................................................................................................................................................
from One Graph to Another 57
Remove a.........................................................................................................................................................
Function from a Graph 57
Show Squares
......................................................................................................................................................... 57
Make a Residual
.........................................................................................................................................................
Plot 58
Plot Values
.........................................................................................................................................................
or Functions with Units 59
Add Grid Lines ..........................................................................................................................................................
to Graph 60
Add Error Bars ..........................................................................................................................................................
to Scatter Plots 60
Filter a Graph.......................................................................................................................................................... 61
Sort Data in a..........................................................................................................................................................
Graph 61
Work with Graph..........................................................................................................................................................
Backgrounds 62
Change the Size..........................................................................................................................................................
of Data Points 63
Change Bar Charts
..........................................................................................................................................................
to Reflect a Function Other Than Count 64
Format Values ..........................................................................................................................................................
displayed in Graph Equations 65
2 Work with Case
...................................................................................................................................
Tables 65
Make a New Case..........................................................................................................................................................
Table 66
Add Cases in ..........................................................................................................................................................
a Case Table 67
Add Attributes..........................................................................................................................................................
in a Case Table 68
Rename an Attribute
..........................................................................................................................................................
in a Case Table 69
Delete Attributes
..........................................................................................................................................................
from a Case Table 69
Add Attributes..........................................................................................................................................................
by Pasting from Somewhere Else 69
Add Attributes..........................................................................................................................................................
from Another Case Table Using Table Join 70
Select Attributes
..........................................................................................................................................................
in a Case Table 71
Select Cases ..........................................................................................................................................................
in a Case Table 72
Work with Formulas
..........................................................................................................................................................
in a Case Table 72
Show/Hide .........................................................................................................................................................
a Case Table’s Formula Row 72
Define an.........................................................................................................................................................
Attribute by Formula When the Formula Row Is Hidden 73
Vertically .........................................................................................................................................................
Resize the Formula Row 73
Adjust Column ..........................................................................................................................................................
Width 74
Sort Data in a..........................................................................................................................................................
Case Table 74
Rearrange Attributes
..........................................................................................................................................................
in a Case Table 74
Hide/Show Attributes
.......................................................................................................................................................... 75
Change the Way ..........................................................................................................................................................
Numbers Are Displayed in a Case Table 75
Manage Units..........................................................................................................................................................
in a Case Table 77
Entering Values
.........................................................................................................................................................
with Units 77
Apply a Unit
.........................................................................................................................................................
Using the Units Row 78
Convert Values
.........................................................................................................................................................
to a Compatible Unit 79
Changing.........................................................................................................................................................
an Attribute Unit Without Converting (Correcting Mistakes) 79
3 Work with Collections
................................................................................................................................... 79
Get Data into ..........................................................................................................................................................
Fathom 80
Paste Data.........................................................................................................................................................
into a Collection 81
Import Data
.........................................................................................................................................................
from a Text File 81
Import U.S.
.........................................................................................................................................................
Census Microdata from IPUMS 82
Change Number.........................................................................................................................................
of Cases Imported from IPUMS 84
Import More Attributes .........................................................................................................................................
from IPUMS 84
Import Households .........................................................................................................................................
from IPUMS 86
Import Specific Geographical
.........................................................................................................................................
Areas from IPUMS 87
Import Historical.........................................................................................................................................
Data from IPUMS 87
Import Data
.........................................................................................................................................................
from the Internet 88
Rename a Collection
.......................................................................................................................................................... 89
Show a Collection’s
..........................................................................................................................................................
Inspector 90
Add an Attribute
..........................................................................................................................................................
to a Collection 91
Add Cases to ..........................................................................................................................................................
a Collection 91
Manage Units..........................................................................................................................................................
in the Collection’s Inspector 92
3
Fathom 2 Help
Define an Attribute
..........................................................................................................................................................
by Formula Using the Collection’s Inspector 93
Generate Random..........................................................................................................................................................
Values 94
Clear a Formula..........................................................................................................................................................
for an Attribute 95
Copy and Paste ..........................................................................................................................................................
an Attribute’s Formula 95
Rename an Attribute
..........................................................................................................................................................
in a Collection’s Inspector 95
Delete Cases .......................................................................................................................................................... 96
Delete an Attribute
..........................................................................................................................................................
in an Inspector 96
Restrict the Cases
..........................................................................................................................................................
You See by Filtering Data 96
Add a Filter
......................................................................................................................................................... 97
Remove a.........................................................................................................................................................
Filter 97
Force a Numeric..........................................................................................................................................................
Attribute to Be Treated Categorically 98
Force a Categorical
..........................................................................................................................................................
Attribute to Be Treated as Numeric 98
Control Behavior
..........................................................................................................................................................
of Categorical Attributes Using Category Sets 99
Create a New
.........................................................................................................................................................
Category Set and Apply It to an Attribute 99
Create a.........................................................................................................................................................
Category Set from Existing Values 100
Reorder .........................................................................................................................................................
Categories in a Category Set 101
Force Nonexistent
.........................................................................................................................................................
Values to Appear in Display Objects 101
Permanently
.........................................................................................................................................................
Force a Numeric Attribute to Be Categorical 102
Remove.........................................................................................................................................................
a Category Set from an Attribute 103
Delete a.........................................................................................................................................................
Category Set 103
Copy a Collection
.......................................................................................................................................................... 103
Merge Two Collections
.......................................................................................................................................................... 104
Copy Cases
.........................................................................................................................................................
from One Collection to Another 104
Prevent Data..........................................................................................................................................................
from Being Changed in Graphs 104
Define a Measure
.......................................................................................................................................................... 105
Restructure ..........................................................................................................................................................
Data-Stack Attributes 105
Export Data ..........................................................................................................................................................
to Other Programs 106
Change the Appearance
..........................................................................................................................................................
of Cases in a Collection 106
4 Work with Summary
...................................................................................................................................
Tables 108
Make a New ..........................................................................................................................................................
Summary Table 108
Add or Remove
..........................................................................................................................................................
a Summary Table’s Attributes 110
Add or Remove
..........................................................................................................................................................
a Summary Table’s Formulas 111
Case Selection
..........................................................................................................................................................
in Summary Tables 111
Use a Summary
..........................................................................................................................................................
Table Without Attributes 112
Sort Categories
..........................................................................................................................................................
in a Summary Table 112
Format Values
..........................................................................................................................................................
in a Summary Table 113
5 Work with Dynamic
...................................................................................................................................
Parameters (Sliders) 113
Change a Slider’s
..........................................................................................................................................................
Scale 114
Animate a Slider
.......................................................................................................................................................... 115
Change a Slider’s
..........................................................................................................................................................
Animation Speed 115
Restrict the Values
..........................................................................................................................................................
a Slider Can Take 116
Limit the Space
..........................................................................................................................................................
Used by Sliders 116
Define a Slider
..........................................................................................................................................................
by Formula 117
Create and Use
..........................................................................................................................................................
Sliders Whose Values Have Units 117
Use a Slider ..........................................................................................................................................................
to Compute Things Based on a Collection 118
Make a Slider..........................................................................................................................................................
Do a Random Walk 119
Make a Non-numeric
..........................................................................................................................................................
Slider and Use It in a Filter 119
Use a Slider ..........................................................................................................................................................
to Create Experiments 120
6 Work with Formulas
................................................................................................................................... 120
Write a Filter.......................................................................................................................................................... 121
Enter a Formula
..........................................................................................................................................................
for an Attribute 121
Work with the ..........................................................................................................................................................
Formula Editor 122
Open the.........................................................................................................................................................
Formula Editor 123
4
Contents
Anatomy.........................................................................................................................................................
of the Formula Editor 125
Use the .........................................................................................................................................................
Keyboard in the Formula Editor 126
Color-Coding
.........................................................................................................................................................
in Formulas 127
Use the .........................................................................................................................................................
Keypad 128
Use the .........................................................................................................................................................
Attribute and Function List 128
Adjust the
.........................................................................................................................................................
Formula Editor Panes 128
Move the.........................................................................................................................................................
Cursor and Select a Portion of an Expression 129
The Effect of Selection
......................................................................................................................................... 129
Make Exponents
......................................................................................................................................................... 129
Make Boolean
.........................................................................................................................................................
Expressions 130
Express .........................................................................................................................................................
a Condition with an If-Statement 130
Compute.........................................................................................................................................................
a Running Sum 131
Recode .........................................................................................................................................................
with a Switch Statement 131
Recode .........................................................................................................................................................
Numeric Values to Categories 132
Recode .........................................................................................................................................................
Ranges of Numeric Values to Categories 134
Use Units.........................................................................................................................................................
in Formulas 134
Frequently Asked
..........................................................................................................................................................
Questions About Formulas 135
7 Work with Surveys
................................................................................................................................... 136
Limitations and
..........................................................................................................................................................
Cautions 137
Use Fathom ..........................................................................................................................................................
Surveys 137
Create a.........................................................................................................................................................
Survey 138
Create Survey .........................................................................................................................................
Instructions 139
Add Multiple-Choice .........................................................................................................................................
Answers 139
Upload a.........................................................................................................................................................
Survey to the Website 140
Make Changes.........................................................................................................................................
to a Survey Form 140
Publish a Survey .........................................................................................................................................
(Allow Students to Take Survey) 141
Publish Survey.........................................................................................................................................
Results (Allow Students to See Survey Results) 142
Upload an Existing .........................................................................................................................................
Fathom Document as a Survey 142
Customize a Survey ......................................................................................................................................... 142
Get a Survey
.........................................................................................................................................................
Filled Out 143
Approve.........................................................................................................................................................
Surveys Created by a Student 143
View and.........................................................................................................................................................
Download the Results 144
Replace Survey.........................................................................................................................................
Data on the Website 144
Delete Survey Results ......................................................................................................................................... 145
Duplicate .........................................................................................................................................................
a Survey 145
Delete a.........................................................................................................................................................
Survey 145
Sort Surveys
......................................................................................................................................................... 145
Search Surveys
......................................................................................................................................................... 146
Install Fathom
.........................................................................................................................................................
Surveys on Additional Computers 146
Manage User..........................................................................................................................................................
Accounts 146
Explanation
.........................................................................................................................................................
of User Roles 147
Log In to.........................................................................................................................................................
Your Fathom Surveys Account 148
Add Instructor
.........................................................................................................................................................
Users 148
Add Student
.........................................................................................................................................................
Users 148
Allow a Student
.........................................................................................................................................................
to Upload Surveys 149
Update User
.........................................................................................................................................................
Information 150
Change .........................................................................................................................................................
Administrators 151
8 Create Experiments
................................................................................................................................... 151
Work with Meters
.......................................................................................................................................................... 151
Make a New
.........................................................................................................................................................
Meter 152
Collect Values
.........................................................................................................................................................
from a Meter 153
Connect Sensors
.......................................................................................................................................................... 153
Set Up an Experiment
.......................................................................................................................................................... 154
5
Fathom 2 Help
Add Attributes
.........................................................................................................................................................
to an Experiment 155
9 Create Simulations
................................................................................................................................... 156
Sampling Simulations
.......................................................................................................................................................... 157
Take a Sample
......................................................................................................................................................... 157
Change .........................................................................................................................................................
the Number of Cases Sampled 158
Collect Another
.........................................................................................................................................................
Sample 158
Collect Measures
..........................................................................................................................................................
and Other Results 159
Define a.........................................................................................................................................................
Measure 159
Make a Measures
.........................................................................................................................................................
Collection 160
Collect More
.........................................................................................................................................................
Measures 160
Create a.........................................................................................................................................................
Collection from the Cells of a Summary Table 161
Collect Results
.........................................................................................................................................................
from Statistical Objects 161
Use Scrambling
..........................................................................................................................................................
to Test for Independence 162
Make a Scrambled
.........................................................................................................................................................
Collection 162
Simulation Examples
.......................................................................................................................................................... 163
Probability
.........................................................................................................................................................
of a Pair 163
Generate.........................................................................................................................................................
a Sampling Distribution 164
Make a Coin-Flip
.........................................................................................................................................................
Simulation 166
Simulate.........................................................................................................................................................
Brownian Motion 168
Test for .........................................................................................................................................................
a Difference in Medians Between Two Groups 170
10 Work with Statistical
...................................................................................................................................
Objects 171
Use Raw Data ..........................................................................................................................................................
in Statistical Inference Objects 172
Enter Summary..........................................................................................................................................................
Statistics in Statistical Inference Objects 172
Shorten the ..........................................................................................................................................................
Output in Statistical Inference Objects 173
Use Sliders to
..........................................................................................................................................................
Vary Summary Information in a Statistical Object 173
Estimate Parameters
..........................................................................................................................................................
and Compute Confidence Intervals 174
Estimate.........................................................................................................................................................
the Population Mean from Raw Data 175
Estimate.........................................................................................................................................................
the Population Mean from Summary Statistics 176
Estimate.........................................................................................................................................................
Difference of Means from Raw Data 176
Estimate.........................................................................................................................................................
Difference of Means from Summary Statistics 178
Estimate.........................................................................................................................................................
the Population Proportion from Raw Data 178
Estimate.........................................................................................................................................................
the Population Proportion from Summary Data 179
Estimate.........................................................................................................................................................
Difference of Proportions from Raw Data 179
Estimate.........................................................................................................................................................
Difference of Proportions from Summary Statistics 182
Test Hypotheses
.......................................................................................................................................................... 183
Using Hypothesis
.........................................................................................................................................................
Test Objects 184
Test Mean.........................................................................................................................................................
from Raw Data (t-Test) 184
Test Mean.........................................................................................................................................................
from Summary Statistics 185
Compare.........................................................................................................................................................
Means from Raw Data (Two-Sample t-Test) 186
Compare.........................................................................................................................................................
Means from Summary Statistics 187
Compare.........................................................................................................................................................
Means When the Values Are Paired (A Paired t-Test) 188
Analysis.........................................................................................................................................................
of Variance 188
Test Proportion
.........................................................................................................................................................
from Summary Data 189
Test Proportion
.........................................................................................................................................................
Against a Value from Raw Data 190
Compare.........................................................................................................................................................
Proportions from Raw Data 191
Compare.........................................................................................................................................................
Proportions from Summary Data 193
Goodness .........................................................................................................................................................
of Fit (Chi-Square) Test from Raw Data 195
Goodness .........................................................................................................................................................
of Fit (Chi-Square Test) from Summary Data: Equal Probabilities 196
Goodness of Fit (Chi-Square Test) from Summary Data: Unequal
Probabilities
......................................................................................................................................................... 198
Test for .........................................................................................................................................................
Independence from Raw Data 199
Test for .........................................................................................................................................................
Independence from Summary Data 200
Test Slope
......................................................................................................................................................... 201
Test Correlation
......................................................................................................................................................... 202
6
Contents
Build a Linear
..........................................................................................................................................................
Model 202
Simple Linear
.........................................................................................................................................................
Regression 203
Multiple .........................................................................................................................................................
Regression 203
Building the Model ......................................................................................................................................... 204
Interpreting the.........................................................................................................................................
Model 206
Adding, Removing, and Moving Predictor Attributes: The Goal of Model
Building ......................................................................................................................................... 207
Changing Multiple .........................................................................................................................................
Regression Output 208
Plotting Predicted .........................................................................................................................................
Values and Residuals 208
Polynomial Regression......................................................................................................................................... 209
Collect Measures
..........................................................................................................................................................
from a Test or Estimate 210
11 Share Your...................................................................................................................................
Results: Presenting and Printing Fathom Results 211
General Techniques
..........................................................................................................................................................
and Ideas 211
Add Comments
.........................................................................................................................................................
to a Collection 212
Work with.........................................................................................................................................................
Text in Text Objects 213
Format Text
.........................................................................................................................................................
in Text Objects 213
Paste Pictures
......................................................................................................................................................... 213
Share Your Fathom
..........................................................................................................................................................
Documents 214
Present in Fathom
.......................................................................................................................................................... 214
Prepare to Print
..........................................................................................................................................................
a Fathom Document 214
Paste Pictures
..........................................................................................................................................................
of Fathom Objects into Another Program 215
215
Part V Movies
1 Graphs ................................................................................................................................... 215
2 Importing Data
................................................................................................................................... 217
3 Inspectors ................................................................................................................................... 218
4 Measures ................................................................................................................................... 219
5 Menus ................................................................................................................................... 219
6 Plotting ................................................................................................................................... 220
7 Summary Tables
................................................................................................................................... 223
8 Surveys ................................................................................................................................... 224
9 Using Formulas
................................................................................................................................... 225
10 Other ................................................................................................................................... 226
227
Part VI Tours
1 Exploring Data—Census
...................................................................................................................................
at Schools 228
2 Data and Prediction—Arm
...................................................................................................................................
Span 234
3 Importing US
...................................................................................................................................
Census Microdata 240
4 Using Formulas
...................................................................................................................................
to Explore the Planets 246
5 Generating...................................................................................................................................
Mathematics—Change Playground 251
6 Simulation—Polling
...................................................................................................................................
Voters 255
7 Testing a Hypothesis—Plant
...................................................................................................................................
Growth 259
8 Testing for...................................................................................................................................
Independence—Pets and Sports 264
9 Numerical Integration—The
...................................................................................................................................
Elevator Experiment 273
10 Classroom...................................................................................................................................
Survey 277
Making the Survey
.......................................................................................................................................................... 278
7
Fathom 2 Help
A First Look..........................................................................................................................................................
at the Survey 279
Modifying the
..........................................................................................................................................................
Survey 280
Taking the Survey
.......................................................................................................................................................... 281
Getting the Data
..........................................................................................................................................................
as the Survey Maker 281
Getting the Data
..........................................................................................................................................................
as a Survey Taker 282
11 Typing Tutor
...................................................................................................................................
Experiment 282
12 How Fast Do
...................................................................................................................................
You Walk?—Measuring Distance with Sensors 283
13 Cooling Water—Measure
...................................................................................................................................
Temperature Over Time 285
14 Timings with
...................................................................................................................................
Photogates 286
289
Part VII Sample Documents
1 Education ................................................................................................................................... 290
2 Fathom Techniques
................................................................................................................................... 291
3 Language and
...................................................................................................................................
the Arts 291
Text Passages
.......................................................................................................................................................... 292
4 Learning Guide
...................................................................................................................................
Starters 293
5 Mathematics
................................................................................................................................... 293
Algebra .......................................................................................................................................................... 294
Calculus .......................................................................................................................................................... 295
Geometry .......................................................................................................................................................... 296
Number Theory
.......................................................................................................................................................... 296
Other Mathematics
.......................................................................................................................................................... 297
Probability .......................................................................................................................................................... 297
6 Science ................................................................................................................................... 299
Astronomy .......................................................................................................................................................... 300
Biology .......................................................................................................................................................... 301
Chemistry and
..........................................................................................................................................................
Physics 301
Earth Sciences
.......................................................................................................................................................... 302
Technology .......................................................................................................................................................... 304
7 Social Science
................................................................................................................................... 304
Canada .......................................................................................................................................................... 304
Canadian.........................................................................................................................................................
Census Files 307
Europe .......................................................................................................................................................... 308
Latin America
.......................................................................................................................................................... 309
United States
.......................................................................................................................................................... 309
States ......................................................................................................................................................... 311
US Census
.........................................................................................................................................................
Files 312
World .......................................................................................................................................................... 313
8 Sports ................................................................................................................................... 314
9 Statistics ................................................................................................................................... 315
Correlation and
..........................................................................................................................................................
Regression 315
Descriptive .......................................................................................................................................................... 317
Distributions.......................................................................................................................................................... 317
Inference .......................................................................................................................................................... 319
Confidence
.........................................................................................................................................................
Intervals 319
Hypothesis
.........................................................................................................................................................
Testing 320
Independence ......................................................................................................................................... 321
Non-parametric.........................................................................................................................................
Tests 322
Modeling .......................................................................................................................................................... 322
8
Contents
328
Part VIII An Overview of Fathom
1 Where Are ...................................................................................................................................
the Data? 328
2 Cases ................................................................................................................................... 329
3 Attributes ................................................................................................................................... 330
Tips for Naming
..........................................................................................................................................................
Attributes 330
Values .......................................................................................................................................................... 330
Numeric vs. ..........................................................................................................................................................
Categorical Attributes 330
Attributes with
..........................................................................................................................................................
Units 331
Attributes with
..........................................................................................................................................................
Category Sets 332
4 Measures ................................................................................................................................... 333
5 Objects and
...................................................................................................................................
Windows 333
6 Collections................................................................................................................................... 334
7 Case Tables
................................................................................................................................... 334
8 Graphs ................................................................................................................................... 335
9 Summary Tables
................................................................................................................................... 335
10 Inspectors ................................................................................................................................... 336
11 Formulas and
...................................................................................................................................
the Formula Editor 338
12 Sliders ................................................................................................................................... 339
13 Meters ................................................................................................................................... 340
14 Derived Collections
................................................................................................................................... 340
15 Statistical Objects
................................................................................................................................... 341
16 Documenting
...................................................................................................................................
Your Work with Text Objects and Pictures 342
342
Part IX Reference
1 Graphs Available
...................................................................................................................................
in Fathom 342
Graphs with..........................................................................................................................................................
Numeric Attributes on One Axis 343
Dot Plot ......................................................................................................................................................... 343
Line Plot......................................................................................................................................................... 344
Histogram......................................................................................................................................................... 345
Ntigram ......................................................................................................................................................... 345
Box Plot......................................................................................................................................................... 346
Percentile
.........................................................................................................................................................
Plot 346
Normal Quantile
.........................................................................................................................................................
Plot 347
Graphs with..........................................................................................................................................................
Categorical Attributes on One Axis 347
Bar Chart......................................................................................................................................................... 348
Ribbon Chart
......................................................................................................................................................... 349
Graphs with..........................................................................................................................................................
Categorical Attributes on Both Axes 350
Split Bar.........................................................................................................................................................
Chart 350
9
Fathom 2 Help
Breakdown
.........................................................................................................................................................
Plot 350
Graphs with..........................................................................................................................................................
Numeric Attributes on Both Axes 351
Scatter Plot
......................................................................................................................................................... 351
Line Scatter
.........................................................................................................................................................
Plot 352
Function.........................................................................................................................................................
Plot 352
2 Inspector Panels
................................................................................................................................... 353
Cases Panel.......................................................................................................................................................... 354
Measures Panel
.......................................................................................................................................................... 355
Comments Panel
.......................................................................................................................................................... 356
Display Panel
.......................................................................................................................................................... 356
Categories Panel
.......................................................................................................................................................... 357
Import URL Panel
.......................................................................................................................................................... 358
Microdata Panel
.......................................................................................................................................................... 358
Sample Panel.......................................................................................................................................................... 359
Collect Measures
..........................................................................................................................................................
Panel 360
Survey Panel.......................................................................................................................................................... 361
Experiment Panel
.......................................................................................................................................................... 362
Scramble Panel
.......................................................................................................................................................... 363
Stack Panel .......................................................................................................................................................... 363
Graph Properties
..........................................................................................................................................................
Panel 364
Graph Cases..........................................................................................................................................................
Panel 365
Slider Properties
..........................................................................................................................................................
Panel 365
Meter Properties
..........................................................................................................................................................
Panel 366
Multiple Regression
..........................................................................................................................................................
Cases Panel 366
3 Fathom Menus
................................................................................................................................... 367
File Menu .......................................................................................................................................................... 368
New ......................................................................................................................................................... 368
Open ......................................................................................................................................................... 368
Open Sample
.........................................................................................................................................................
Document 369
Close ......................................................................................................................................................... 369
Save ......................................................................................................................................................... 369
Save As......................................................................................................................................................... 369
Import ......................................................................................................................................................... 370
Import From File ......................................................................................................................................... 370
Import From URL ......................................................................................................................................... 370
Import U.S. Census .........................................................................................................................................
Data 371
Export Collection
......................................................................................................................................................... 371
Revert Collection
......................................................................................................................................................... 371
Show Page.........................................................................................................................................................
Breaks 372
Print Preview
.........................................................................................................................................................
(Win only) 372
Page Setup/Print
.........................................................................................................................................................
Setup 372
Print ......................................................................................................................................................... 372
Exit/Quit......................................................................................................................................................... 372
Edit Menu .......................................................................................................................................................... 372
Undo ......................................................................................................................................................... 373
Redo ......................................................................................................................................................... 373
Cut ......................................................................................................................................................... 374
Copy ......................................................................................................................................................... 374
Copy As.........................................................................................................................................................
Picture 374
Paste ......................................................................................................................................................... 374
Delete ......................................................................................................................................................... 375
Select All
......................................................................................................................................................... 375
Edit Formula
......................................................................................................................................................... 375
10
Contents
Cut Formula
......................................................................................................................................................... 376
Copy Formula
......................................................................................................................................................... 376
Paste Formula
......................................................................................................................................................... 376
Clear Formula
......................................................................................................................................................... 376
Text Font......................................................................................................................................................... 376
Text Size......................................................................................................................................................... 376
Text Style
......................................................................................................................................................... 377
Show/Hide.........................................................................................................................................................
Text Palette 377
Preferences
......................................................................................................................................................... 377
Object Menu.......................................................................................................................................................... 378
New ......................................................................................................................................................... 378
Inspect [Object]
......................................................................................................................................................... 378
Duplicate.........................................................................................................................................................
[Object] 379
Delete [Object]
......................................................................................................................................................... 379
Add/Remove
.........................................................................................................................................................
Filter 379
Hide [Object]
......................................................................................................................................................... 380
Show Hidden
.........................................................................................................................................................
Object(s) 380
View In Window
......................................................................................................................................................... 380
Collection Menu
.......................................................................................................................................................... 380
Rerandomize/Sample
.........................................................................................................................................................
More Cases/Scramble Again/Collect More Measures 381
New Cases......................................................................................................................................................... 382
Prevent .........................................................................................................................................................
Changing Values in Graphs 382
Rename.........................................................................................................................................................
Collection 382
Sample .........................................................................................................................................................
Cases 382
Scramble .........................................................................................................................................................
Attribute Values 383
Collect Measures
......................................................................................................................................................... 383
Stack Attributes
......................................................................................................................................................... 384
Table Menu .......................................................................................................................................................... 385
New Attribute
......................................................................................................................................................... 385
Rename.........................................................................................................................................................
Attribute 385
Hide Attribute(s)
......................................................................................................................................................... 385
Show Hidden
.........................................................................................................................................................
Attributes 386
Format Attribute
......................................................................................................................................................... 386
Autofit Column
.........................................................................................................................................................
Widths 387
Show/Hide.........................................................................................................................................................
Units 388
Show/Hide.........................................................................................................................................................
Formulas 388
Sort Ascending
......................................................................................................................................................... 389
Sort Descending
......................................................................................................................................................... 389
Use As Caption
......................................................................................................................................................... 389
Graph Menu.......................................................................................................................................................... 390
Remove.........................................................................................................................................................
X Attribute: [AttributeName] 391
Remove.........................................................................................................................................................
Y Attribute: [AttributeName] 391
Remove.........................................................................................................................................................
Legend Attribute: [AttributeName] 391
Show/Hide.........................................................................................................................................................
Axis Links 391
Unlink [X/Y]
.........................................................................................................................................................
Axis 392
Rescale .........................................................................................................................................................
Graph Axes 392
Stack Dots
......................................................................................................................................................... 392
Scale ......................................................................................................................................................... 393
Least-Squares
.........................................................................................................................................................
Line 394
Median-Median
.........................................................................................................................................................
Line 394
What Is a Median-Median
.........................................................................................................................................
Line? 395
Add Movable
.........................................................................................................................................................
Line 395
Remove.........................................................................................................................................................
Movable Line 395
Show Squares
......................................................................................................................................................... 396
11
Fathom 2 Help
Make Residual
.........................................................................................................................................................
Plot 396
Lock Intercept
.........................................................................................................................................................
at Zero 397
Plot Value
......................................................................................................................................................... 397
Plot Function
......................................................................................................................................................... 398
Sort Bars......................................................................................................................................................... 399
Sort By Plotted
.........................................................................................................................................................
Value 399
Format Value
......................................................................................................................................................... 400
Slider Menu .......................................................................................................................................................... 400
Format Value
......................................................................................................................................................... 400
Summary Menu .......................................................................................................................................................... 400
Add Formula
......................................................................................................................................................... 401
Add Basic.........................................................................................................................................................
Statistics 402
Add Five-Number
.........................................................................................................................................................
Summary 402
Remove.........................................................................................................................................................
Attribute 404
Sort Categories
.........................................................................................................................................................
by Formula Value 404
Create Collection
.........................................................................................................................................................
From Cells 404
Format Value
......................................................................................................................................................... 405
Estimate Menu.......................................................................................................................................................... 405
Verbose......................................................................................................................................................... 405
Remove.........................................................................................................................................................
Attribute 406
Collect Results
.........................................................................................................................................................
As Measures 406
Test Menu .......................................................................................................................................................... 407
Verbose......................................................................................................................................................... 407
Remove.........................................................................................................................................................
Attribute 408
Collect Results
.........................................................................................................................................................
As Measures 408
Show Test.........................................................................................................................................................
Statistic Distribution/Show p_hat Distribution 409
Model Menu.......................................................................................................................................................... 409
Verbose......................................................................................................................................................... 410
Remove.........................................................................................................................................................
Attribute 410
Collect Results
.........................................................................................................................................................
As Measures 411
Hide/Show.........................................................................................................................................................
Sequential Contributions Chart 411
Hide/Show.........................................................................................................................................................
ANOVA Table 412
Show/Hide.........................................................................................................................................................
Predictors in ANOVA Table 412
Window Menu .......................................................................................................................................................... 413
Cascade.........................................................................................................................................................
(Win only) 413
[Documents]
......................................................................................................................................................... 413
Help Menu .......................................................................................................................................................... 413
Context (Right-Click)
..........................................................................................................................................................
Menus 413
4 Fathom Operators,
...................................................................................................................................
Functions, and Units 413
Operators .......................................................................................................................................................... 414
Arithmetic Functions
.......................................................................................................................................................... 415
Trigonometric..........................................................................................................................................................
Functions 416
Statistics Functions
.......................................................................................................................................................... 417
One Attribute
.........................................................................................................................................................
Statistics Functions 417
Transformations
......................................................................................................................................................... 419
Two Attributes
......................................................................................................................................................... 420
Random Functions
.......................................................................................................................................................... 421
Date/Time Functions
.......................................................................................................................................................... 422
Distribution ..........................................................................................................................................................
Functions 423
Binomial.........................................................................................................................................................
Distribution 425
Chi-Square
.........................................................................................................................................................
Distribution 425
Exponential
.........................................................................................................................................................
Distribution 426
F-Distribution
......................................................................................................................................................... 427
Geometric.........................................................................................................................................................
Distribution 428
12
Contents
Normal Distribution
......................................................................................................................................................... 428
Other Distributions
......................................................................................................................................................... 429
List Functions
.......................................................................................................................................................... 429
Logical Functions
.......................................................................................................................................................... 430
Conditional Functions
.......................................................................................................................................................... 431
Text Functions
.......................................................................................................................................................... 431
Other .......................................................................................................................................................... 433
Special Values
.......................................................................................................................................................... 434
Special Values
..........................................................................................................................................................
for Summary Table 434
Special Value
..........................................................................................................................................................
for Histogram 435
Special Value
..........................................................................................................................................................
for Ntigram 435
Units .......................................................................................................................................................... 435
5 Keyboard Shortcuts
................................................................................................................................... 441
444
Part X Troubleshooting
1 Problems with
...................................................................................................................................
Data 445
My Data Are ..........................................................................................................................................................
Gone 445
My Data Don’t
..........................................................................................................................................................
Paste into Fathom 446
Fathom Won’t..........................................................................................................................................................
Let Me Open/Import My File 446
A Web Page ..........................................................................................................................................................
Doesn’t Import 446
Census Microdata
..........................................................................................................................................................
Import Attempt Failed 446
Fathom Treats
..........................................................................................................................................................
Numeric Data as Categorical 447
I Assigned a..........................................................................................................................................................
Wrong Unit and Can’t Fix It 447
I’m Having Trouble
..........................................................................................................................................................
Opening a Fathom Document from My Browser 448
2 Problems with
...................................................................................................................................
Graphs 448
I Want to Split
..........................................................................................................................................................
a Numeric Graph, Not Make a Scatter Plot 448
I Have Frequencies:
..........................................................................................................................................................
How Do I Make a Histogram? 449
I Can’t Drag ..........................................................................................................................................................
Data in a Graph 449
The Text in My
..........................................................................................................................................................
Graph Is Unreadable or Takes Up Too Much Space 449
The Dots Became
..........................................................................................................................................................
Colored Squares 450
No Data Show ..........................................................................................................................................................
in the Graph 450
Problems with..........................................................................................................................................................
Residual Plots 451
3 Problems with
...................................................................................................................................
Sliders 451
I Can’t Drag ..........................................................................................................................................................
the Slider’s Thumb 451
My Slider Rounds
..........................................................................................................................................................
Its Value How Can I Make It Stop? 451
4 Problems with
...................................................................................................................................
Inspectors 451
I Can’t See the
..........................................................................................................................................................
Panel I Need 451
Where’s the ..........................................................................................................................................................
Show Details Button? 452
5 Problems with
...................................................................................................................................
Formulas 452
Fathom Won’t
..........................................................................................................................................................
Recognize a Function I Typed 452
Error Messages
..........................................................................................................................................................
for Formulas 452
6 Problems with
...................................................................................................................................
Statistical Inference Objects 452
I Don’t See Statistical
..........................................................................................................................................................
Objects on the Shelf 453
This Object Says
..........................................................................................................................................................
I’m Required to Use Raw Data, But I Should Be Able to Enter Summary Data 453
7 Problems with
...................................................................................................................................
Surveys 453
I Can't Find the
..........................................................................................................................................................
Fathom Surveys Website 453
I Made a Survey
..........................................................................................................................................................
and Can’t Upload It 453
I Uploaded a..........................................................................................................................................................
Survey, but Don't See It on the Website for Students 453
8 Problems with
...................................................................................................................................
Experiments 453
My Sensor Is..........................................................................................................................................................
Not Working 454
13
Fathom 2 Help
My Experiment
..........................................................................................................................................................
Is Running Very Slowly 454
9 My Problem
...................................................................................................................................
Still Isn’t Solved 454
454
Part XI Glossary
Index 461
14
Fathom 2 Help
1 Fathom 2 Help
Getting Started 23
How To’s 33
Movies 215
Watch short movies showing how to accomplish specific basic tasks.
Reference 342
Read comprehensive reference materials, including descriptions of all the graphs Fathom makes; menu-
by-menu explanations of menu commands; explanations of each of Fathom’s inspectors; the lists of
operators, functions, and units; or keyboard shortcuts.
Troubleshooting 444
See how to solve a specific problem you’ve encountered.
KCP Technologies
15
Fathom 2 Help
16
Fathom 2 Help
It has been our good fortune to collaborate with the IPUMS-USA project, part of the Minnesota
Population Center, to make possible Fathom’s census microdata import capability.
Citation: Steven Ruggles, Matthew Sobek, Trent Alexander, Catherine A. Fitch, Ronald Goeken, Patricia
Kelly Hall, Miriam King, Colin C. Davis, and Chad Ronnander. Integrated Public Use Microdata Series:
Version 3.0 [Machine-readable database]. Minneapolis, MN: Minnesota Population Center [producer and
distributor], 2004.
Acknowledgements
Fathom Version 2 would not have come to be without the help and support of a wonderful community
of educators. We gratefully acknowledge Luis Acuna, Jim Albert, Christian Aviles-Scott, Ron Avitzur,
Wesley Bailey, Rolf Biehler, Bill Block, Jim Bohan, Frank Briody, Gail Burrill, Ruth Carver, Karen Coe,
Jere Confrey, Gord Cooke, Brian Cooley, Steward Craven, Gretchen Davis, Colin Davis, Bob delMas,
Chris deMers, Lew Douglas, Brian Erwin, Len Fellman, Joan Garfield, Rob Gould, Jim Hammerman,
Chris Hancock, David Hoppe, Cliff Konold, Katie Makar, Craig Miller, Paul Myers, Sara Nerlove, John
Olive, Chris Olsen, Henri Picciotto, Andee Rubin, Steve Ruggles, Al Sagel, Mike Shaughnessy, Franklin
Sheehan, Tom Steinke, Tony Thrall, Carolyn Wales, Joel Yan, Joshua Zucker, and Steven Zucker.
No project this size comes to fruition without the help of many family members and friends. Among
these are Brigid McCaw, Lauren Finzer, Devin Finzer, Liam, Kieran, and Ramona Swenson,
Anne Erickson, and Meg Holmberg.
Excerpts from Version 1 Acknowledgements
17
Fathom 2 Help
Nick Jackiw not only contributed in major ways to Fathom’s design, he continually brought us back to
the fundamental vision of educational software we are pursuing.
Date/Time
Fathom now understands date and time formats for data and sliders.
Units
Fathom 2 recognizes and understands units, from pounds to fathoms. This means you can now have data
that answer the question “How many what’s?” Unit designations appear on graphs and other objects. You
can easily convert from one unit to another; formulaic computations respect units and can include explicit
mention of units.
See Manage Units in a Case Table 77 or Manage Units in the Collection’s Inspector 92 .
See Units 435 for a comprehensive list of built-in units Fathom recognizes.
18
What's New in Fathom 2?
Colors
Colors are used to tie corresponding elements together. For example, both the case table and the Cases
panel have the same blue shading. Computed values show a gray background. Empty cells have a yellow
background. Attribute names, measure names, functions, units, slider names, and constants each have a
unique color that is used wherever that kind of word appears.
See Color-Coding in Formulas 127 .
Link Axes
You can now link axes so that their scales match, and stay matched, even when rescaled.
See Dynamically Link Axes to Each Other 43 .
Graph Properties
Now graphs have their own inspectors that allow you to change the size of points and axis bounds, add
error bars, determine whether axes will automatically rescale, and even change the direction of increasing
numbers.
See Graph Properties Panel 364 .
New Bar Charts
19
Fathom 2 Help
Two bar charts, one with side-by-side bars and the other with bars split into multiple groups, are now
possible.
See Bar Chart 348 and Split Bar Chart 350 .
Show Squares to Plotted Functions
In Fathom 1 the squares of residuals could only be shown for lines. Now they can be shown for plotted
functions. The sum of squares is shown, as well.
See Show Squares 57 .
Multiple Movable Lines in Bivariate Plots
Now you can add any number of movable lines to a bivariate plot.
Least-Squares Regression and Median-Median Lines Computed Separately for Each Cell
When a scatter plot has been split with a legend attribute, or with multiple attributes on an axis,
computing a fitted line shows the line and its equation for each cell rather than a single line for all the
data.
See Least-Squares Line 394 and Median-Median Line 394 .
Plotted Function Tracing
As your mouse moves over a line or plotted function in a bivariate graph, a red dot appears indicating
where you are, and the coordinates of the dot appear in the status bar. By clicking the mouse, you can
display the coordinates directly in the graph.
See Trace on a Function or Fitted Line 56 .
Powerful New Slider Capabilities
Fathom 2 greatly increases slider usefulness while preserving accessibility for getting started with sliders.
You can now control animation speed, restrict values to multiples, compute slider value with a formula,
refer to collection attributes and measures in a slider formula, and allow self-reference in slider formulas.
See Work with Dynamic Parameters (Sliders) 113 or Slider Properties Panel 365 .
Linear Models
The statistical objects include a linear model object as well as interval estimates and hypothesis tests.
See Build a Linear Model 202 .
Multiple Linear Regression Model
This new statistical object brings important statistical capabilities to Fathom’s dynamic dragging
environment. It has its own set of attributes for working with predicted and residual values.
See Multiple Regression 203 .
Correlation Matrix
Add numeric attributes to both dimensions of a summary table, and you get a flexible correlation matrix.
20
What's New in Fathom 2?
21
Fathom 2 Help
22
What's New in Fathom 2?
When the cursor is in a denominator or exponent, a tip appears informing the user that the right-arrow
key will move the cursor out of the denominator or exponent.
Rerandomization Is Much Less Aggressive
Under many, but not all, circumstances in which Fathom 1 would rerandomize all computed values,
Version 2 will only rerandomize directly affected values.
New Functions
missing itemSum monthName
isPrime itemCount now
linRegrPredicted firstItem second
linRegrResidual lastItem stringToDateTime
logRelativeError nthItem stringToTimeDuration
scalar sortItems ticks
unitOf date today
cardIcon dayOfMonth year
indexOfCategory dayOfWeek
lookupValueByIndex dayOfWeekName
lookupValueByKey dayOfYear
popCovariance hour
sampleCovariance minute
month
Other Improvements
You can now copy a picture of an object (especially a graph) even when cases are selected.
Double-clicking a column boundary in a case table automatically resizes that column to be the correct
width to show all its values. A menu command allows you to do this to multiple columns or all
columns.
Clicking on cells in a summary table selects all the cases corresponding to that cell. If all the cases
corresponding to a cell are selected, the cell highlights.
The command File | Open Sample Document takes you directly into the Sample Documents folder
so you don’t have to search for it (provided your sample documents are stored in the same folder as the
application).
The Help menu has commands that will open your Web browser to useful online locations.
On the Macintosh, z+? will launch Help.
3 Getting Started
People have many different ways of approaching a new program, so we have provided a variety of ways
to learn Fathom.
The subtopics in this section show just the basics of getting and graphing some data. They also provide
links to the rest of Fathom Help for more functionality. To start, go to Getting Data 24 .
Another option is to start with Getting Started, a Fathom document with data in it and instructions for
starting to look at that data.
If you like taking step-by-step guided tours, you’ll find them in the Learning Guide that comes with
23
Fathom 2 Help
Fathom. These can also be accessed from the Tours 227 section of Help, which includes tours for the
survey and experiment features (these tours are not in the printed book). Printable PDF files of the
Learning Guide and Fathom Help are located in the Documentation folder.
If you want to see brief movies showing Fathom in use, go to Movies 215 .
If you like to understand the logic of a program before using it, see An Overview of Fathom 328 first.
I want to use some data already in Fathom: Open a Fathom Sample Document 24
I want to type data to work with (Hint: use a case table): Enter Data from Scratch 25
I have data in electronic form that I want to bring into Fathom (Hint: paste it into a collection): Copy
and Paste Data from Another Program 26
I want to work with data about individual people from the U.S. Census: Import U.S. Census Microdata
30
I want to begin by simulating data (take a brief tour and flip a virtual coin): Have Fathom Randomly
Generate Data 27
Some other ways in which you can get data from outside sources are surveys and sensors.
I want to use web-based surveys to gather responses to questions as data (see Using Fathom Surveys)
136 .
I want to collect values from Vernier sensors and construct experiments around them (see Create
Experiments 151 and Connect Sensors 153 ).
24
Getting Started
In Sample Documents are folders with documents for learning about Fathom (for example, Learning
Guide Starters contains documents you need to follow the guided tours, and Fathom Techniques
shows some of Fathom’s advanced features), but most folders, such as Science or Sports, are subjects
or topics. You could pick an interesting area and explore.
See Sample Documents 289 to read about all of Fathom’s sample documents, or choose one of our
recommended documents.
2. Navigate to the document you want to open.
Recommendations:
Science | Elements.ftm
Science | RegionalCarbon.ftm
Social Science | World | CountriesUnemployment.ftm
Education | United States | FacultySalaries.ftm
3. Double-click the file to open it.
In order to graph the data in your sample document, you need to be able to see the data’s attributes
(some programs call them variables), to drop them on graph axes. If you can’t see the attribute names,
double-click the file’s collection, the box of gold balls, to show its inspector.
You are now ready to go to Beginner Graphing, or watch a movie showing how to make graphs: Four
Basic Objects-Case Table, Graph, and Summary Table.
25
Fathom 2 Help
1. Drag a case table from the object shelf, and drop it in the
document.
In a case table, rows represent cases and columns represent
attributes (or variables).
2. Click <new>, type a name for the new attribute and press
Enter. (If necessary, repeat this step to add more attributes;
pressing Tab instead of Enter moves you to the next column.)
3. Click in the blank cell under the attribute name and begin typing values. (Press Tab to move from cell
to cell.)
When you named your first attribute, Fathom created an empty collection to hold your data (a little,
empty box). The collection is where your data are actually stored. Deleting the collection deletes your
data. When you add cases by typing values, the collection icon fills with gold balls.
Collection 1 Collection 1
26
Getting Started
If your data didn’t come in perfectly, you might need to fix things (for example, give the attributes better
names or delete a bogus case).
If what you got is too much trouble to fix, you might make the data conform to Fathom’s structure (as in
the screen shot above) in its current program and try again. Or you could try importing it, rather than
copy/paste. See Import Data from a Text File 81 or Import Data from the Internet 88 for more detail.
You are ready to graph your data, go to Beginner Graphing 31 .
See also
Rename an Attribute in a Case Table 69
27
Fathom 2 Help
The first attribute has been named flip, and the cursor is in
place to name the second attribute. Notice that the value cell
has a different colored background to show there’s no value in
it. You’ll tell Fathom how to generate values with a formula.
4. Double-click the Formula cell for flip.
28
Getting Started
Fathom has accepted the formula, but you don’t yet have a
value because you don’t have any cases.
7. Select the collection (the empty box icon) by clicking it,
and choose Collection | New Cases.
8. Type 10 and press Enter or Return.
You’ve just created a fair coin and flipped it ten times. You
can scroll through your ten flips by clicking the arrows in the
bottom-left corner of the inspector.
Choosing Rerandomize from the Collection menu will
generate a new set of flips.
Here are a few ideas for where to go from here (some are simple; others require delving into other areas
of Fathom Help).
Add a second attribute with the formula runLength(flip) to compute how many heads and tails
you get in a row.
Graph your data. (Drag a graph from the shelf and drop an attribute on its horizontal axis.
See Beginner Graphing 31 for more detail.)
Look at all of the cases at once in a case table. (Select the box of gold balls and drag a case
table into your document. See Work with Case Tables 65 for more detail.)
Add more cases.
Edit the formula to make the coin unfair, for example, randomPick("heads", "tails", "tails", "tails")
then see if you get longer or shorter run lengths.
Open the collection by dragging a corner, and change the images of the cases to those of
heads and tails of a nickel using formulas in the inspector’s Display panel. (see Change the
Appearance of Cases in a Collection 106 .)
Try some of Fathom’s random number generators, which can be found in the formula editor’s list of
functions. You can also learn about them by reading the Random Functions 421 section of Help.
See also
Work with Summary Tables 108
Create Simulations 156
29
Fathom 2 Help
1. In a new, empty Fathom document, choose File | Import | U.S. Census Data.
Fathom creates an empty collection and opens its inspector to the Microdata panel.
Fathom provides default selections for you. You can download the request at this point or expand the
headings to specify sample size, places and years to sample from (Choosing Cases), and more information
about each person (Attributes). You can also decide to import samples of households, rather than of
individual people. (see the Subsections below for more detail.)
2. Click Download Data in the lower-right corner of the inspector.
Fathom sends a request to the IPUMS Web site and returns the requested data. When the download is
complete, the status area changes to reflect the number of cases imported, and the collection icon image
changes to a box full of gold balls.
The categorical attributes (see Numeric vs. Categorical Attributes 330 ) and Census_Year have category
sets assigned to them automatically. (see Control Behavior of Categorical Attributes Using Category Sets
99 .)
Subsections
Change Number of Cases Imported from IPUMS 84
30
Getting Started
See also
Work with Graphs 33
2. Drag an attribute from the case table or inspector, and drop it on the prompt in the
graph under the horizontal axis or to the left of the vertical axis.
31
Fathom 2 Help
You get a graph. If the attribute you graphed was numeric, you get a dot plot 343 ; if it was categorical,
you get a bar chart.
Change the graph’s type using the pop-up menu at top right in the graph.
You can add an attribute to the same axis, to the other axis, or in the plot area. See Replace an Attribute
on a Graph with Another Attribute 35 , Add Another Attribute to an Already Occupied Axis 35 , or Add
an Attribute to the Middle of a Graph 36 .
You can have as many graphs as you want. If you select data in one graph, it will be selected everywhere.
See Select Data in a Graph 37 .
With just some data and graphs, you can go far in data exploration and analysis. Of course, there’s much
more you can do in Fathom, as described in the rest of Fathom Help.
See also
Work with Graphs 33
32
How To’s
4 How To’s
This section lists many common Fathom tasks and provides directions for each. If you’re new to Fathom
and want to dive right in, consider starting with the Getting Started 23 section of Help.
Subsections
Work with Graphs 33
Work with Case Tables 65
Work with Collections 79
Work with Summary Tables 108
Work with Dynamic Parameters (Sliders) 113
Work with Formulas 120
Work with Surveys 136
Create Experiments 151
Create Simulations 156
Work with Statistical Objects 171
Share Your Results: Presenting and Printing Fathom Results 211
1. Make a new (empty) graph by dragging one from the shelf into a blank area of your document or by
choosing Object | New | Graph.
2. Drag the name of an attribute to one of the axes of the graph.
The first sequence shows graphing a numeric attribute, and the second shows graphing a categorical
attribute.
33
Fathom 2 Help
The prompt that appears in the graphs tells you how to force Fathom to treat your numeric attribute as
categorical or to ignore categorical values and plot only numbers in a categorical attribute. Most of the
time, you just drop what you want to see, so you can ignore the prompt. (For more detail, see Force a
Categorical Attribute to Be Treated as Numeric 98 and Control Behavior of Categorical Attributes Using
Category Sets 99 .)
This produces a graph of the data, though it may not be the kind you want.
3. Choose the kind of graph you want from the
pop-up menu on the graph.
34
How To’s
Graph with Another Attribute 35 or Add Another Attribute to an Already Occupied Axis 35
Dropping an attribute on the plus sign of an axis will add the attribute to the one already there, rather than
replace it.
See also
Add Another Attribute to an Already Occupied Axis 35
Drag an attribute to the graph, and drop it directly on top of the plus sign. (When you’re over the right
spot, the whole-axis highlighting disappears, and highlighting appears around the plus.)
35
Fathom 2 Help
Fathom adds the attribute to the graph, giving you a split dot plot (or histogram, box plot, etc.).
Here are circumstances in which you can add attributes to already occupied axes:
Add as many numeric attributes as you want to a numeric univariate plot (such as a dot plot
or histogram).
Add as many numeric attributes as you want to one axis of a scatter plot.
Add one categorical attribute to a bar chart.
See also
Replace an Attribute on a Graph with Another Attribute 35
Dropping sex in the plot area of the dot plot of age changes the dots representing the men from gray
circles to blue squares (each category gets its own symbol). A legend appears below the plot area.
Split categorical plot (by dropping a second categorical attribute in the middle of a bar or ribbon chart)
Dropping sex in the plot area of the ribbon chart of Labor_force_status divides each section of ribbon to
36
How To’s
reflect the proportion of each sex within that labor status category. A legend appears below the plot area.
You can drop a numeric attribute in the plot area of a dot plot or scatter plot.
In the scatter plot of mass and speed of extra solar planets, period is dropped in the plot area. Each case
is now colored to reflect its period, and a legend appears below the plot area.
Remove an attribute from the graph by selecting the graph and choosing Graph | Remove
[AttributeLocation]: [AttributeName].
To remove an attribute from a graph, select the graph by clicking on it and choose Graph | Remove
[location:] [AttributeName].
If you want to replace an attribute on an axis or in a legend with another attribute, you don’t need to
remove it first; simply drop the attribute(s) you want on top of the attribute you want to replace (see
Replace an Attribute on a Graph with Another Attribute 35 ).
37
Fathom 2 Help
To select multiple bars or dots, draw a selection rectangle around them. All cases, bars, or
bins even partially within the rectangle will be selected.
To select all cases in a category, click the legend symbol for that category.
38
How To’s
39
Fathom 2 Help
In the first graph, you see the movement caused by the Kobe earthquake over its entire duration. In the
second graph, after zooming in to a time slice of three minutes, the graph axes scale automatically to
show ticks in terms of seconds.
To watch movies illustrating some of these techniques, see Rescaling Axes and Zooming in Graphs.
40
How To’s
Note: You can always return to the original all-points state by choosing Graph | Rescale Graph Axes or
by re-selecting the plot type from the pop-up menu in the graph.
See also
Change Axes Numerically 41
Zoom In or Out in a Graph 41
Prevent Graph Axes from Automatically Rescaling 45
1. Double-click in the graph to show its inspector; if necessary, go to the Properties panel.
2. Edit the value for the upper or lower bounds of the x- or y-axis as desired (Tab moves you to the next
property), and press Enter or Return.
If you want an axis to stay put and not automatically rescale, you can change that axis’s AutoRescale
property to false.
See also
Change Axes by Dragging 40
Zoom In or Out in a Graph 41
Graph Properties Panel 364
1. Move the cursor over a graph axis or plot area, and hold down Ctrl (Win) Option (Mac). The cursor
changes to a zoom-in cursor.
2. Click on an axis to expand the graph around that point or drag a rectangle around the area you want to
focus on (the rectangle expands, filling the plot area).
Holding down Shift as well reverses the process, making you zoom out.
41
Fathom 2 Help
Return to the original all-points state by choosing Graph | Rescale Graph Axes or by re-
selecting the plot type from the pop-up menu in the graph.
You can see how it works in the movie Zooming in Graphs.
See also
Change Axis Scales 40
Whether changing the order of categories on a graph axis affects that graph only, or all objects in which that
attribute is displayed depends on whether the attribute has a category set applied to it or not. If no
category set, then the effect is on one object only. If there is a category set, then the effect is on all
objects. (see Control Behavior of Categorical Attributes Using Category Sets 99 .)
See also
Attributes with Category Sets 332
Sometimes you want to reverse a graph’s axis; for example, when looking at data about the stars,
constellations often appear backward in graphs, as in this graph of Pleiades.
1. Double-click the graph to show its inspector; if necessary, go to the Properties panel.
42
How To’s
Each axis has a property allowing you to reverse its scale, which is set to false by default.
2. Set the [axis]ReverseScale property for the axis you want by selecting true from the popup.
The horizontal (x) axis has been reversed; Pleiades is oriented correctly.
43
Fathom 2 Help
An animation shows that the axes are now linked, and the chain link is closed.
Rescaling either axis will automatically rescale the other to keep them matched up.
44
How To’s
When cases are added to a collection, Fathom checks the bounds of axes of graphs of the data in that
collection to make sure all the data will show on the graph. Usually this is a good thing, but there are
times when you would like the bounds of the graph’s axes to remain fixed while cases are added, for
example, if you are collecting measures to build a sampling distribution, and you want to watch the graph
as it grows.
1. Double-click the graph, to show its inspector (if necessary, go to the Properties panel).
Each axis of the inspected graph has a property for auto rescaling. By default, these are set to true.
2. Set these to false for each axis you want to set.
Fathom will not change that axis’s bounds automatically, but you can change it manually either by
dragging or by choosing Graph | Rescale Graph Axes.
45
Fathom 2 Help
When looking at a histogram or ntigram, you may find it informative to see how the graph looks with
different bin sizes. There are two ways to accomplish this: dynamically, by dragging a bin edge, or
numerically, by typing a bin width (histogram) or bin number (ntigram) in the Properties panel of its
inspector. There’s a movie that covers both methods A Dot Plot, a Histogram, and a Box Plot.
1. Double-click the graph, to show its inspector (if necessary, go to the Properties panel).
2. Change the value for the binWidth property. (For ntigrams, change the numBins property.)
See also
Change Bin Width by Dragging 46
46
How To’s
47
Fathom 2 Help
If you want to specify a bin width for your histogram numerically (or the number of bins in an ntigram
), use the graph’s inspector.
1. Double-click the graph to show its inspector; if necessary, go to the Properties panel.
2. Change the value in the binWidth property, and press Enter or Return.
See also
Change Bin Width by Dragging 40
Subsections
Add a Movable Line to a Graph 48
Plot a Line at a Given Value on a Graph 49
48
How To’s
You can add an unfitted, movable line to any graph that has at least one numeric attribute.
1. Select the graph.
2. Choose Graph | Add Movable Line.
If the graph has numeric attributes on both axes, you get an oblique line, and its equation appears below
the plot. Dragging that line by its middle keeps the slope but changes the intercept (translates the line);
dragging by either end changes the slope (rotates the line). The cursor you get when you are over the line
changes to tell you which kind of drag you’re about to perform. (Watch the movie Movable Lines in
Scatter Plots to see how to move this line.)
If the graph has numeric attributes on only one axis, you get a line perpendicular to that axis, and the
value of that line appears below the plot area. Drag the line to move it, and the value updates dynamically.
You can add more movable lines by choosing Graph | Add Movable Line again.
Remove a movable line by selecting it (click it or its equation) and choosing Graph | Remove
Movable Line.
See also
Show Squares 57
49
Fathom 2 Help
The formula result appears as a colored line and as a color-coded value at the bottom of the plot.
Note: If you leave the argument of a function blank, Fathom will use the attribute on the axis as the
function’s argument. You could, instead, use the name of another attribute in the formula.
You can plot as many values as you want.
If the graph is split, a line appears for each subgroup; the value reported below is for all the data in the
graph.
50
How To’s
See also
Show Measures of Spread in a Graph 51
To show measures of spread in a graph (such as standard deviation), you don’t really want to simply plot
the standard deviation. Suppose, for example, that your data range from 100 to 200, with a standard
deviation of 30. The value wouldn’t appear on the graph, even if it were meaningful to see it there. What
you want to do is to plot, say, mean plus and minus 1 standard deviation. Here’s how:
1. Select the graph.
2. Choose Graph | Plot Value.
7. Click OK or press Enter or Return to accept the formula and close the formula editor.
See also
Add Fitted Lines to a Graph 53
51
Fathom 2 Help
If you have a histogram and you want to see how well it approximates a given distribution, you can plot
a function directly on the graph over the histogram bars.
1. Select the graph.
2. Choose Graph | Scale | Density.
This changes the graph’s axis from a frequency scale to a density scale. (The total area of the bins is 1.)
3. Choose Graph | Plot Function.
4. Enter the formula for the distribution. (You can see what distributions are available and what
arguments they require by opening the Functions list in the formula editor. Instead of typing numeric
values for parameters, you can use functions for the data (such as mean) that get their values from the
data in the graph.)
52
How To’s
You can plot two different kinds of fitted lines on a scatter plot: the least-squares line and the median-
median line (see What Is a Median-Median Line? 395 ).
1. Select the graph.
2. Choose Graph | Least-Squares Line or Graph | Median-Median Line.
The corresponding line appears on the graph, and its equation appears below the plot area. Choose the
command again to remove the line.
You can force a least-squares line to go through the origin by choosing Graph | Lock
Intercept at Zero.
Note: The least-squares and median-median lines update dynamically as the data change. (Drag a data
point to see whether it influences the line.)
When you have a multiattribute scatter plot (more than one attribute on a given axis), you get a line and
equation for each.
See also
Show Squares 57
Make a Residual Plot 396
Trace on a Function or Fitted Line 56
53
Fathom 2 Help
If you aren’t sure exactly what to plot, you can use one or more sliders in the formula and change the
sliders to change the curve dynamically. See Plot a Function You Can Vary Dynamically (Using a Slider)
55 .
See also
Make a Residual Plot 396
Show Squares 57
Trace on a Function or Fitted Line 56
Remove a Function from a Graph 57
4.1.11.7 Plot Functions on a Function Plot
Although specifically designed for working with data, Fathom’s graph objects can also be used as
function plots. These plots are especially powerful when the functions are governed by one or more
sliders.
1. Make a new, empty graph by dragging one from the shelf or choosing Object | New | Graph.
2. From the graph’s pop-up menu, choose Function Plot.
54
How To’s
3. Choose Graph | Plot Function to show the formula editor, and enter the function you want to plot.
4. Click OK or press Enter or Return to accept the formula and close the formula editor.
By making one or more sliders and using their names in your functions, you make the functions dynamic.
Drag the slider and watch the function change. See Plot a Function You Can Vary Dynamically (Using a
Slider) 55 .
You can add as many functions (or values) as you want to a graph.
Dropping an attribute on a function plot will remove everything plotted on it.
1. Make a slider by dragging one from the shelf or choosing Object | New | Slider.
2. Give the slider a meaningful name (such as denominator).
3. Select the graph.
4. Choose Graph | Plot Function.
This shows a formula editor.
5. Define the function you want plotted, and click OK or press Enter or Return to accept the formula
and close the formula editor.
If you are already editing a function when you decide to use a slider, you can put the slider’s (future) name
in the formula and make the slider after closing the formula editor. (Your graph will have a #Name Not
55
Fathom 2 Help
Recognized# error that goes away once you’ve created and named a slider.)
See also
Show Squares 57
Make a Residual Plot 396
Trace on a Function or Fitted Line 56
2. When this dot is visible, hold down the mouse button; the dot will trace the function and its
coordinates will appear in the plot area.
Because clicking on a movable line selects it for dragging, this feature is not available for movable lines.
Instead, you can plot a static line (using Plot Function) and trace that.
56
How To’s
You can copy a function from one graph and paste it onto another.
1. Select the function you want to copy (click either the curve itself or its formula).
2. Choose Edit | Copy Formula.
3. Select the receiving graph and choose Graph | Plot Function.
4. Press Ctrl+V (Win) Option+V (Mac) and then Enter or Return to accept the formula and close the
formula editor.
Alternatively, you could duplicate the graph with the function by selecting it and choosing
Object | Duplicate Graph.
Fathom constructs vertical segments from every point to the line and then makes those segments into
sides of squares. Fathom also displays the sum of the areas of those squares.
57
Fathom 2 Help
Show Squares also works with the movable line in a univariate plot. This can be used to show that the
mean is a least-squares value for univariate data.
You can also use this feature in conjunction with a residual plot.
The residual plot is always about half the height of the main plot.
Note: The vertical scale on the residual plot may be very different from the scale in the original. Residual
plots can help you see features that are invisible on the original graph.
If you have a graph with more than one thing plotted on it, you can change the focus of the residual plot
by selecting a different equation or line.
58
How To’s
When you want to plot a value on a graph with units, you must include the units for any numbers in the
formula. For example, if you were to try to plot the value 5 on this graph, no line would show because 5
doesn’t have units. But the value 5s will show just fine because it has the right units.
Fathom is slightly more forgiving with plotted functions than plotted values. For example, an attempt
to plot 0.1 Time on the graph at right shows the line even though the values have units of seconds instead
of meters. But the color of the plotted function is magenta, the color Fathom uses for invalid values, and
59
Fathom 2 Help
60
How To’s
61
Fathom 2 Help
If a category set is applied to the attribute, dragging or sorting categories will change the order of its
categories; all objects displaying the attribute will use this new sort order.
When you have a split numeric graph with a value plotted on it, you can sort the groups by that value.
1. Select the graph by clicking on it.
2. Choose Graph | Sort By Plotted Value.
If you have more than one value plotted on the graph, first select one by clicking on its expression below
the plot, then sort.
See also
Control Behavior of Categorical Attributes Using Category Sets 99
Add a Background
62
How To’s
Note: The image will not resize with the plot unless you lock
the background.
Lock a Background
To scale the background image with the graph, lock the background to the graph scales.
1. Select the graph.
2. Choose Graph | Lock Background to Graph Axes.
3. Scaling or zooming in the graph will now change the image with the data.
Delete a Background
To remove the background from a graph,
1. Select the graph.
2. Make sure that no cases are selected.
3. Choose Edit | Cut Background to remove the background and place it in the clipboard, or Edit |
Delete Background to remove the background altogether.
You can control the size of data points for any graph that
shows data as dots (dot plots, box plots with outliers, normal
quantile plots, scatter plots, and breakdown plots).
1. Double-click an axis or the plot area of the graph to show
its inspector; if necessary, go to the Properties panel by
clicking its tab.
2. Change the value of the property pointSizeMultiplier (which
is 5 by default).
63
Fathom 2 Help
The figures above show a graph with the default point size of 5, changed to 3, then to 7.
Point size is in pixels. You may need to experiment to get the look you want.
You can also control the size of data points using a formula that depends on other attribute values.
1. Go to the graph inspector's Cases panel.
2. Click in the formula column of the property pointSize.
3. Enter the formula that will determine the point size. This formula will use the base point size specified
in the property pointSizeMultiplier.
If you want to make everything in the document bigger (all fonts and other elements in all objects), for
giving a presentation projected onto a screen, for example, you should change the document preference
to use a larger font size in objects. See Preferences 377 .
64
How To’s
You can now format the values displayed in the graph equations. The functionality is similar to that of
the Format Attribute command from the Table menu (see Change the Way Numbers Are Displayed in
a Case Table 75 )
65
Fathom 2 Help
Subsections
Make a New Case Table 66
Add Attributes by Pasting from Somewhere Else 69
Add Attributes from Another Case Table Using Table Join 70
If you have data to enter into Fathom, start by making an empty case table. If you have data and want to
see or manipulate them, you want to connect the table to the collection. To make an empty case table,
drag one from the object shelf at the top of the Fathom window, and drop it in your document.
To make a case table of data you already have, select the collection, then drag a case table from the shelf
and drop it in the document.
66
How To’s
If you forget to select the collection first, drop the collection’s name in the body of the case table.
See also
Add Attributes in a Case Table 68
67
Fathom 2 Help
This method won’t work if all your attributes are generated by formula, because you can’t edit calculated
values. In that instance, you need the New Cases command.
1. Select the case table (or its collection).
2. Choose Collection | New Cases.
You get a dialog box that will add one case by default.
3. If necessary, type the number of cases you want to add and click OK.To insert cases before the end of
the collection (for example, if you have time series data and you left out a case):
Select a case in the case table, choose Collection | New Cases, specify how many cases to
add in the dialog box, and click OK.
Added cases are inserted before the selected case.
If you are about to paste data, you don’t need to create the cases first; Fathom creates them when you
paste. If you are about to import data, you don’t even need to make a case table or collection. Fathom
creates the collection when you import.
See also
Get Data into Fathom 80
Work with Formulas in a Case Table 72
2. Type the attribute name and press Enter or Return on your keyboard to accept the attribute name. If
you have more than one attribute to add, press Tab to move to the next column.
Note: Attribute names must begin with a letter and can contain only letters, numbers, and underscores (_).
See Tips for Naming Attributes 330 for more detail. If you type an invalid name, Fathom warns you of
your error and removes invalid characters. If you don’t want to keep the name Fathom gives you, double-
click the name to change it.
To add an attribute before the last attribute:
1. Select the attribute to the right of where you want to add the new attribute.
68
How To’s
If you want to add attributes, but don’t have a case table of the collection, you can also Add an Attribute
to a Collection 91 using the collection’s inspector.
See also
Work with Formulas in a Case Table 72
69
Fathom 2 Help
If an attribute is selected in the case table, pasting attributes will insert them to the left of the selected
attribute.
If this method doesn’t work right for some reason, consider pasting all of your data in a spreadsheet or
word-processing program to get them set up in a Fathomable configuration, then pasting them into a new
Fathom collection or importing them (see Paste Data into a Collection 81 or Import Data from a Text
File 81 ).
4.2.7 Add Attributes from Another Case Table Using Table Join
Suppose you have two collections of data about your class, and you want to graph attributes from them
both on the same graph. If both collections have an attribute that uniquely identifies each student in the
class, you can easily bring the missing attributes from one collection into another. The cases don’t have
to be in the same order in the two collections. Fathom uses the "key" attributes to match cases properly.
This matching operation is called a table join.
1. Put both collections in the same document, each with its own case table. (Copy the collection from
one document, and paste it into the document you want to add to. Make case tables for both
collections; see Make a New Case Table 66 .)
2. Drag the name of the key attribute from the source collection, and drop it on the name of the key
attribute in the destination collection. (Below, student is dropped on name.)
70
How To’s
our class
nam e s ibs pe ts w o r k fave Clas s <ne w >
1 Sharon B. 1 no no PE
2 Anu 3 yes no Math
3 Noel 0 no no
4 Aneesa 2 no no Eng
5 Mario 2 yes no Eng
6 Aaron 3 yes no PE
7 Lerena 1 no yes Math
8 Elizabeth 1 yes no Eng
9 Sharon T. 0 yes yes Hist
10 Jasmine 4 no no Hist
Fathom adds any attributes in the source collection that are missing from the destination to the attributes
in the destination collection. The added attribute(s) have formulas using the Lookup function (see
lookupValueByKey 434 ). If you want to delete the source collection, you first need to clear the formulas
(otherwise, the data you’ve just brought over will be lost).
See also
Add Attributes by Pasting from Somewhere Else 69
71
Fathom 2 Help
When you select a case, it is highlighted in all the graphs and tables in which it appears. This is useful
when you want to see how some cases lie in relation to others. Also, to delete cases from a collection, you
must first select them.
Select a single case by clicking its row number.
Select more than one case by dragging along the cases’ row numbers.
Select contiguous cases by clicking the row number at one end of the desired section. Then
press Shift and click on the row number of the case at the other end of the series.
To select cases that are not contiguous, press Ctrl (Win) z(Mac) and click each row number.
Select all cases displayed in the table by selecting the case table and choosing Edit | Select All
Cases.
See also
Select Data in a Graph 37
Case Selection in Summary Tables 111
72
How To’s
The formula row (labeled with an equals sign) now appears in the case table. The formulas defining
attributes appear; the formula cells for noncomputed attributes are shaded to indicate that they have no
formula.
1. Double-click the formula cell for an attribute to enter or edit its formula.
2. Choose Table | Hide Formulas to hide the formula row.
Can’t see your formula? If you know a formula is there, but you can’t see it, or can’t see all of it, resize the
formula row height, the attribute width, or both. See Adjust Column Width 74 and Vertically Resize the
Formula Row. 73
4.2.10.2 Define an Attribute by Formula When the Formula Row Is Hidden
You don’t need to show the formula row to enter or edit a formula in the case table.
1. Select the attribute whose formula you want to edit.
2. Choose Edit | Edit Formula to show the formula editor.
3. Enter or edit your formula and click OK or press Enter or Return to accept the formula and close the
editor.
To edit an attribute’s formula, double-click one of its cells. (Cells defined by formula have a gray
background.)
If entering the formula wiped out your data, immediately choose Undo Edit Formula.
See also
Work with the Formula Editor 122
Fathom Operators, Functions, and Units 413
1. Move the cursor over the lower border of the formula row until it becomes the resize cursor.
2. Drag the border down and release.
73
Fathom 2 Help
See also
Select Attributes in a Case Table 71
Drag the attribute’s name to the border between the attributes where you want to move it.
74
How To’s
See also
Select Attributes in a Case Table 71
1. Select the attributes you want to hide. (To select more than one, hold down the Shift key and click.)
2. Choose Table | Hide Attribute.
To show previously hidden attributes:
1. Select the case table.
2. Choose Table | Show Hidden Attributes.
See also
Select Attributes in a Case Table 71
Delete Attributes from a Case Table 69
1. Select the attribute whose number format you want to change by clicking on its name in a case table.
2. Choose Table | Number Format.
Dynamic
75
Fathom 2 Help
The default numeric format, Dynamic, is a good general-purpose format. It displays up to six significant
digits with a decimal point, unless the number is too big or too small, in which case it displays scientific
notation with six significant digits. It doesn’t display trailing zeroes after the decimal point.
Fixed Decimal
All numbers will have the same number of digits after the decimal point, the number specified in the
Decimal Places field.
Scientific Notation
A number expressed in scientific notation has one digit to the left of the decimal point, as many digits as
are specified in the Significant Digits field, and an exponent that specifies by what power of 10 to
multiply.
Engineering Notation
In engineering notation, the number of digits to the left of the decimal point is either one, two, or three,
adjusted so that the power of 10 will always be a multiple of 3.
This table gives some examples, all with six digits specified in the Number Format dialog box.
76
How To’s
See also
Manage Units in the Collection’s Inspector 92
Units 435
77
Fathom 2 Help
space between the numeric part of the value and the unit. (You can type that space or not, but Fathom
will add it if it isn’t there.)
If some values are in a different, but compatible, unit, Fathom will automatically convert them to the
attribute’s unit. As soon as a value with a unit that differs from the default is accepted (when you leave
the cell), Fathom converts the value to the value of the default unit.
Values that have a unit incompatible with the unit assigned to the attribute are colored magenta to set
them off as invalid. These values will be ignored by other Fathom objects, such as graphs and statistical
objects.
To use a unit that isn’t built-in, you need to enter it into the units row of a case table or the units field in
an inspector (see Manage Units in the Collection’s Inspector 92 ). (This feature is handy if you have
measured things in invented units. For example, suppose the class decides to have a student named Edna
measure car lengths using her arm span as the unit; you can create a unit called ednas.)
You can change an attribute’s default unit or assign or remove a unit in the case table’s units row, which
you can show by choosing Table | Show Units. (You can also do these things in a collection inspector’s
Details pane.)
See also
Manage Units in the Collection’s Inspector 92
1. If the units row is not already showing, show it by choosing Table | Show Units (the case table must
be selected to enable the Table menu).
2. Type the unit’s name, or a recognized abbreviation, in the unit cell for the attribute to which you want
to assign the unit.
3. Press Enter or Return.
The long form of the unit appears in the unit cell, and each value has the short form of the unit appended
to it.
If you apply a unit Fathom doesn’t recognize, Fathom responds with a dialog box asking whether or not
you want to use your unit. You can invent new units, but Fathom will not automatically convert them (do
conversions by defining a formula for a new attribute).
For a list of built-in units and recognized abbreviations, see Units 435 . Watch a movie about Working with
Units.
78
How To’s
You can use the units row to convert values from one unit to another compatible unit. (If you don’t have
a case table, you can also Manage Units in the Collection’s Inspector 92 .)
1. If the units row isn’t showing, choose Table | Show Units.
2. Select the unit you want to change by clicking the unit cell.
3. Type the new unit.
4. Press Enter or Return.
Fathom will convert the values into the new units.
See also
Changing an Attribute Unit Without Converting (Correcting Mistakes) 79
Suppose you realize that the units assigned to a certain attribute are completely wrong. The numbers are
right, but the units need to be changed without the numbers changing. Perhaps someone got confused, as
shown at right, and used seconds as an angle measure. Or perhaps someone thought all the values were
inches, when, in fact, they were centimeters.
1. Delete the unit from the units row and press Return or Enter.
Notice that the units of the values have gone away but the numbers have stayed the same.
2. Type in the desired units and press Return or Enter.
See also
Manage Units in the Collection’s Inspector 92
79
Fathom 2 Help
represented by a gold ball (see Change the Appearance of Cases in a Collection 106 ).
This section of Fathom Help addresses how to do things with the collection. Many tasks in this section
(such as Add an Attribute to a Collection 91 , Define an Attribute by Formula Using the Collection’s
Inspector 93 , Copy and Paste an Attribute’s Formula 95 , Clear a Formula for an Attribute 95 , Rename
an Attribute in a Collection’s Inspector 95 , Delete an Attribute 96 , or Manage Units in the Collection’s
Inspector 92 ) could also be done in a case table.
Many of these tasks are done using the collection’s inspector (such as Control Behavior of Categorical
Attributes Using Category Sets 99 , or Define a Measure 105 ). See Show a Collection’s Inspector 90 .
To show a collection’s inspector, double-click it, or select it and choose Object | Inspect
Collection.
Other Subsections
Add Cases to a Collection 91
Generate Random Values 94
Delete Cases 96
Restrict the Cases You See by Filtering Data 96
Force a Numeric Attribute to Be Treated Categorically 98
Force a Categorical Attribute to Be Treated as Numeric 98
Merge Two Collections 104
Prevent Data from Being Changed in Graphs 104
Restructure Data—Stack Attributes 105
Export Data to Other Programs 106
See also
Derived Collections 340
Collection Menu 380
Inspector Panels 353
80
How To’s
census data from the Integrated Public Use Microdata Series (IPUMS) Web site at the University of
Minnesota; see Import U.S. Census Microdata from IPUMS 30 .
To type data, use a case table (see Enter Data from Scratch 25 for detailed instructions).
You can also Generate Random Values 94 . See Random Functions 421 for the list of random functions.
Have Fathom Randomly Generate Data 27 is an introductory tutorial on coin flipping. Also see
Simulation Examples 163 .
Fathom allows you to paste cases you’ve copied from somewhere else into a collection. For this to work,
your data need to be arranged correctly: columns represent attributes (or variables); each row is a case (or
record). It’s best if the first row consists of valid attribute names, which begin with a letter or underscore,
and include nothing but letters, underscores, and numbers (no spaces or other characters, such as percent
signs). If the data you want aren’t at all as described, pasting might not work; try exporting and importing
instead.
1. Open the file with your data in its current program.
2. Copy the data (including the attribute names).
3. Switch to Fathom.
In Fathom, a data set is stored in a collection, so you need to make a collection to hold the copied data.
If this method of getting data doesn’t work, maybe you could Import Data from a Text File 81 instead,
or consider pasting your data into a word processor or spreadsheet where you can prepare the collection
to be copied. If the data is on a Web page, you could Import Data from the Internet 88 by dropping the
URL from your browser into Fathom.
See also
Where Are the Data? 328
81
Fathom 2 Help
1. Export the data from the program as a tab-delimited text file, including the attribute names (if
possible). (see the program’s documentation if you need help with this; most programs that work with
data have a tab-delimited text file export option.)
2. Switch to Fathom.
3. Choose File | Import | Import From File.
4. Select the file you want to import and click Open.
5. Check that the import worked as expected by making a case table (with the collection selected, drag a
case table from the shelf).
If your data imports but the first case is used as the attribute names, you can fix it in Fathom. (Double-
click an attribute name to edit it. After fixing the attribute names, delete the first case.) To fix the data
before importing, open the text file in a word processing or spreadsheet program and add a line at the top
with the attribute names. Then import this edited file.
If the import didn’t work well enough (that is, if it would be too much work to make it right), consider
working with the data in a spreadsheet or text-editing program to get it into “Fathomable” shape.
Spreadsheets, for example, often have a command that turns “Cases as Columns” into “Cases as Rows.”
See also
Make a New Case Table 66
Paste Data into a Collection 81
Import Data from the Internet 88
82
How To’s
1. In a new, empty Fathom document, choose File | Import | U.S. Census Data.
Fathom creates an empty collection and opens its inspector to the Microdata panel.
Fathom provides default selections for you. You can download the request at this point or expand the
headings to specify sample size, places and years to sample from (Choosing Cases), and more information
about each person (Attributes). You can also decide to import samples of households, rather than of
individual people. (see the Subsections below for more detail.)
2. Click Download Data in the lower-right corner of the inspector.
Fathom sends a request to the IPUMS Web site and returns the requested data. When the download is
complete, the status area changes to reflect the number of cases imported, and the collection icon image
changes to a box full of gold balls.
The categorical attributes (see Numeric vs. Categorical Attributes 330 ) and Census_Year have category
sets assigned to them automatically. (see Control Behavior of Categorical Attributes Using Category Sets
99 .)
Subsections
Change Number of Cases Imported from IPUMS 84
See also
Work with Graphs 33
83
Fathom 2 Help
By default, you get a sample of 500 cases. You can change your sample size to be between 1 and 10,000
cases.
1. In the Microdata panel of the inspector, expand the Choosing Cases list.
2. Select Sample Size.
3. Edit the Maximum Sample Size. (You can’t import more than 10,000 cases at a time.)
Some samples (such as those from small states) may not have as many cases as you’ve specified. This field
doesn’t determine the number of cases; rather, it sets a limit.
See also
Import More Attributes from IPUMS 84
Import Households from IPUMS 86
Import Specific Geographical Areas from IPUMS 87
Fathom offers a few dozen of the attributes available through IPUMS. The attributes fall into three
categories: Year and Location, Person, and Household.
84
How To’s
Person attributes, such as age and total personal income, apply to individual people. Household
attributes, such as number of bedrooms or monthly rent, apply to entire households. Each case has the
household information repeated. If you want to analyze household information (such as proportion of
households that are farms), you should import household records, not people records.
Within the Person and Household lists, attributes have been grouped further.
1. Select an item in the list on the left to show the attributes available in that category.
2. Click an attribute to select it for import.
The attribute is added to the request (as can be seen in the list of currently requested attributes).
Not all attributes are available for all years. (The Census Bureau didn’t always ask all of the questions it
does now. Names are available only for data from 1920 or earlier.) You can see what years an attribute is
available for by pointing at it with the cursor and reading Fathom’s status area in the lower left of the
Fathom window.
See also
Change Number of Cases Imported from IPUMS 84
85
Fathom 2 Help
By default, the Microdata panel is set to request samples of individual people. You can instead import
samples of households. When importing households, the attributes that apply to individuals (such as age
and marital status) don’t import.
To change your data request to Households:
1. If necessary, open the Choosing Cases list.
2. Click People/Households to bring up the pane with these choices.
3. Click Households Only.
4. Open the Attributes list then the Household list to choose household attributes.
Household attributes are grouped into three sections: Basic (owned or rented), Economic (rent and
mortgage), and Dwelling (number of bedrooms and age of residence).
See also
Change Number of Cases Imported from IPUMS 84
86
How To’s
The default IPUMS request samples people or households from all 50 states and Washington, D.C. You
can narrow the geography by choosing specific states or metropolitan areas. Don’t choose a combination
of states and metro areas (for example, Anchorage and Alaska) or the import won’t work.
1. If necessary, open the Choosing Cases list.
2. Select either States or Metropolitan Areas to show the list of available states or cities.
3. Check the desired places.
The request now contains the checked items.
Tip: If you select several states in the Choosing Cases section of the Microdata panel, you’ll want the
attribute State, so you can tell which case came from which state.
When the data are downloaded, the people or households will all have been sampled from the requested
area(s).
Metropolitan area samples are not available for 1960.
See also
Change Number of Cases Imported from IPUMS 84
The IPUMS collection includes samples from 1850 to 2000. The year 1890 is missing; unfortunately, the
records were destroyed by fire. 1930 is also missing.
87
Fathom 2 Help
See also
Change Number of Cases Imported from IPUMS 84
88
How To’s
Fathom cannot import data in PDF format or in the form of images. The page you import needs to be
the one that has the data on it.
Drag the URL icon from your browser into an open Fathom document. Fathom imports the
data and makes a new collection. (There is a movie illustrating this: Internet Import by
Dragging the URL.)
Or…
1. Copy the URL from the address field in your browser.
2. Choose File | Import | Import From URL. A dialog box appears.
3. Type or paste the URL of the data you want into the box and click OK. Fathom imports the data and
makes a new collection.
After importing, look at the data in a case table to see whether it came in properly. Often you have to do
a little fixing before the collection is right.
Select the collection and drag a case table from the shelf into your document.
You might also want to check the collection’s comments. If Fathom did not put important information
into the comments, you may want to copy and paste comments separately. (Comments can go in the
Comments panel of the collection’s inspector, see Comments Panel 356 .)
Not all Web pages are in a format Fathom can decipher. You could paste your data in a word processing
or spreadsheet program to clean it up, then copy and paste it into Fathom.
You can find links to some “Fathomable” Web pages by following the Links to Data link on the Fathom
Resource Center, or go there directly by clicking here: Links to Data.
See also
Make a New Case Table 66
Paste Data into a Collection 81
Import Data from a Text File 81
Import U.S. Census Microdata from IPUMS 30
89
Fathom 2 Help
Double-click the collection name, type the new name into the dialog box, and click OK.
Or …
Select a collection, graph, or case table and choose Collection | Rename Collection.
Click a tab along the top of the inspector to go to another panel. (The attribute names are in
the Cases panel.)
See also
Inspectors 336
Inspect [Object] 378
Inspector Panels 353
90
How To’s
To add an attribute to the end of the collection, you can use either an inspector (on the Cases panel)
or a case table. To add an attribute anywhere else, you need a case table.
1. Show the collection’s inspector by double-clicking the collection; if necessary, go to the Cases panel
(the leftmost panel).
2. Click <new>, type a name, and press Enter or Return.
91
Fathom 2 Help
Delete Cases 96
You don’t need a case table to apply, change, or remove units from attributes. The Details pane of the
Cases panel of a collection’s inspector includes a Units field.
1. In the Cases panel of the collection’s inspector, if Details aren’t showing, click Show Details.
A new pane appears, below the attribute names.
2. Select an attribute to show the Units field for that attribute.
3. Edit the unit for the selected attribute in the Units field.
In the illustration, typing cm in the Units field applies centimeters to length.
To convert an attribute’s values, select the current unit and replace it with the unit you want to
convert to.
To change a unit without converting its values (usually done to correct a mistake), first delete
the current unit and press Enter or Return to clear the unit, then type the new unit. If you
don’t clear the mistaken unit first, Fathom will try to convert.
See also
Manage Units in a Case Table 77
Units 435
Attributes with Units 331
92
How To’s
There are many reasons to create attributes defined by formula. Maybe you are basing an attribute on
other attributes (such as multiplying height times width to compute areas), generating random values, or
recoding data.
1. In the Cases panel of the collection’s inspector, double-click the attribute’s formula cell to show the
formula editor.
2. Enter the formula telling Fathom how to compute the attribute’s values and press Enter or Return to
accept the formula and close the formula editor.
Warning: Defining a formula for an attribute that already has values will replace those values with the
formula’s results. To compute values based on an attribute, add a new attribute and define it by formula.
See also
Work with Formulas in a Case Table 65
Work with Formulas 120
Fathom Operators, Functions, and Units 413
93
Fathom 2 Help
Fathom has many built-in functions for generating random values. You can read about them in Random
Functions 421 , or in the help pane of the formula editor (open the list of functions, then the list of
random functions; click on a specific function and information about it appears in the bottom of the
formula editor).
1. Create an attribute in the Cases panel of the collection’s inspector.
2. Double-click the attribute’s formula cell to show the formula editor.
3. Enter a random number function and whatever arguments it requires (if any), and press Enter or
Return to accept the formula and close the formula editor.
94
How To’s
If no value appears, you probably need to add one or more cases to the collection.
4. Choose Collection | New Cases, specify the number of cases, and click OK.
Tip: You can use functions and existing attributes as arguments in formulas. For example, if you want
random numbers from a distribution with a mean and standard deviation of some attribute already in the
collection, you can use mean(attributeName) and StdDev(attributeName)as arguments in the randomNormal
function.
For more examples of randomly generated data, see Have Fathom Randomly Generate Data 27 ,
Probability of a Pair 163 , Make a Coin-Flip Simulation 166 , and Simulate Brownian Motion 168 .
1. In the Cases panel of the collection’s inspector, select the attribute whose formula you want to clear.
2. Choose Edit | Clear Formula.
95
Fathom 2 Help
1. Select the cases you want to delete (in a graph, case table, or open collection).
2. Choose Edit | Delete Cases.
See also
Delete Attributes from a Case Table 69
See also
Add a Filter 97
96
How To’s
Delete Cases 96
When you want to see only a subset of your collection’s data, you can add a filter to the collection or any
object connected with it. A filter applied to the collection filters every object connected to it (graphs,
tables, statistical objects); a filter applied to any other object filters only that object.
1. Select the object you want to filter and choose Object | Add Filter.
2. Enter the expression for the filter in the formula editor and click OK.
Appropriate expressions include:
sex = "female"
(age ≥ 21) and (age ≤ 65)
(color = "red") or (color = "green")
The formula for the filter appears on the bottom of the filtered object.
A filtered collection still shows all its cases, but the captions of the filtered-out cases are gray instead of
black. A graph or case table attached to a filtered collection will show only the cases that pass through the
filter.
Two graphs attached to the same collection can have different filters. This can be useful when you want
to compare the results of a display or analysis under two different conditions.
Edit the formula for a filter by double-clicking it.
See also
Restrict the Cases You See by Filtering Data 96
Delete Cases 96
97
Fathom 2 Help
Add a Filter 97
Delete Cases 96
Hold down Shift as you drop the attribute onto the Fathom object (graph, summary table, or statistical
object). You can start the drag normally and press Shift just before the drop.
If you want the attribute to always be treated as categorical, generate a category set from it (see
Attributes with Category Sets 332 ).
1. In the Cases panel of the collection inspector, click Show Details.
2. Select the attribute and choose Generate From Values from the Set pop-up menu.
Fathom generates a category set from the attribute’s values. Even if all values are numeric, if you apply a
category set to an attribute, it will always be treated as categorical when dropped on Fathom objects.
In this example, the class was divided into groups to conduct an experiment; to get a dot plot of distance
split by group (rather than a scatter plot), group needs to be interpreted as categorical.
See also
Force a Categorical Attribute to Be Treated as Numeric 98
Control Behavior of Categorical Attributes Using Category Sets 99
Hold down Ctrl (Win) Option (Mac) key as you drop the attribute onto the object.
Warning: This method has disadvantages: You have to do it each time you drop the attribute, and, more
seriously, values that have both numeric and non-numeric values will be ignored completely. For example,
98
How To’s
many attributes have a “top end” value, such as 7+ representing women who have had seven or more
children. Leaving these values out of your analysis distorts your results. Instead, you should make an
attribute that strips out the non-numeric characters.
Add a new attribute and give it the formula StringToNumber(mixedAttribute).
The new attribute will have only the numeric values stripped of any non-numeric characters. You could
clear the formula at this point. You should document (in the collection’s Comments or a text object) that
the “top end” value represents “that value and above.”
See also
stringToNumber 433
1. Show the collection inspector by double-clicking the collection or by selecting the collection or a case
table and choosing Object | Inspect Collection.
2. Click Categories to go to that panel.
3. Click <new> and name the category set. (Press Enter or Return to create more than one.)
99
Fathom 2 Help
4. Click in the Categories cell for a set, and type the categories, in the order you want them to appear in
graphs and such, separated by commas.
This creates the sets; now they need to be applied to attributes.
5. Go to the inspector’s Cases panel.
6. Click Show Details. (If the button says Hide Details, then details are already showing; the button
toggles between the two states.)
7. Select one or more attributes (Shift+click), and choose a set name from the Set pop-up menu.
Here, the set indicating level of interest is applied to four attributes.
100
How To’s
If you have an attribute that’s already populated with values and you want to generate a category set from
them, you can do this in the collection inspector’s Cases panel.
1. Show the collection inspector by selecting the collection or a case table and choosing Object |
Inspect Collection.
2. Click the Categories tab to go to that panel.
3. Click <new>, type a name for the category set, and press Enter or Return.
101
Fathom 2 Help
4. Click in the Categories cell for that set, and type the categories, in the order you want them to appear
in graphs and such, separated by commas.
5. Go to the Cases panel.
6. Click Show Details. (If the button says Hide Details, then details are already showing; the button
toggles between the two states.)
7. Select the attribute to which you want to apply the category set, and choose the name of the set from
the Set pop-up menu. (You can apply a set to more than one attribute at a time.)
The attribute now has the category set applied to it. When the attribute is dropped on an object, all the
categories will appear (whether or not any cases have them), in the order determined in the category set
definition you created. Values that do not match those categories (if there are any) will be colored pink in
case tables and inspectors and will appear under an <invalid> category in other objects.
See also
Reorder Categories in a Category Set 101
Remove a Category Set from an Attribute 103
Delete a Category Set 103
102
How To’s
If an attribute has a category set assigned to it and you don’t like the way it affects the attribute, you can
remove the assignment.
1. Show the collection inspector, and go to the Cases panel. (Select either the collection or a case table
for it, and choose Object | Inspect Collection; if necessary, click the Cases tab).
2. Click Show Details.
3. Select the attribute whose category set you want to remove, and choose None from the Set pop-up
menu.
The set has been removed, and the attribute will revert to its default behavior. The category set still exists
and can be applied to other attributes.
See also
Control Behavior of Categorical Attributes Using Category Sets 99
To get rid of a category set entirely (as opposed to removing it from an attribute):
103
Fathom 2 Help
104
How To’s
If you have defined a measure to build a sampling distribution, you can collect measures by dropping a
measure’s name on an empty collection. Each measure becomes an attribute in the measures collection.
See also
Collect Measures 383
Measures 333
Create Simulations 156
Work with the Formula Editor 122
The problem is that Fathom treats each row as a case, but, in fact, the two numbers in each row are not
logically part of the same case. It would be better to work with a collection in which each case represents
one person.
105
Fathom 2 Help
This creates a new collection called, in this case, Stacked heights, in which the values for all the columns
of attributes are stacked on top of each other and a new attribute, called Group, is created with values
corresponding to the names of the original columns. You will probably want to rename the attributes.
See also
Stack Panel 363
106
How To’s
The Display panel for the Planets collection (see Display Panel 356 )
x and y control the horizontal and vertical placement of the center of the case icon (with (0, 0)
being the upper left).
image controls the appearance of the case icon (a gold ball by default).
width and height control the icon’s size.
caption controls what text, if any, appears under each case in the open collection, as well as
what appears in the status bar of the Fathom window when you point at a data point in a
graph.
Although you can type values directly for each case, you usually want them governed by formula.
Planets Rerandomize
Numeric attributes are in pixels. In the inspector above, notice that width and height are set to the radii of
each planet and that the x and y attributes, in turn, depend on them.
The modulo function is often handy for the x attribute when you want cases ordered in rows
and columns. By using sliders in these formulas, you can have cases move around in the open
collection when the slider’s value is changed or when the slider is animated. See Work with
Dynamic Parameters (Sliders) 113 .
107
Fathom 2 Help
You don’t need to change values or formulas manually for x and y; you can simply drag case
icons around in the open collection to change their values.
You can use the switch function for image to vary images; in this example, it is done using
caseIndex, but more often, it will be done using cases’ data. (For example, a switch on an
attribute for die rolls can make dice appear with the face that reflects the value of the roll, or
an if-statement can give you a nickel’s heads or tails.)
Fathom has dozens of built-in icons that can be accessed by formula in the formula editor’s
functions list under Icon Names. The sample document FathomIconList.ftm shows them all.
You can copy an image from somewhere else, select a case in an open collection, and choose
Edit | Paste As Case Icon to give cases any image you want.
To have cases appear without captions, use the formula " " (empty quotation marks).
The concat function allows you to use more than one attribute and other text as captions. For
example, in a collection with two attributes, Name and Pet, you could write a formula that
produces captions such as “Lauren has a parakeet” using this formula:
concat(Name, " has a ", Pet)
You can have Fathom create a caption formula that concatenates many attributes by selecting
the attributes in a case table and choosing Table | Use As Caption.
The Sample Documents folder includes many files that use the open collection to display
data.
108
How To’s
1. Make an empty summary table by dragging one from the shelf or choosing Object | New | Summary
Table.
2. Drop one or more attributes either to the right of or below the prompt. (If necessary, Show a
Collection’s Inspector 90 by double-clicking the collection.)
no data
Drop an attribute here
When you drag an attribute over a summary table, arrows appear around the prompt. You can drop the
attribute on the prompt (one of the arrows will highlight, indicating where the attribute will be put), or on
the column header to the right of the prompt, or on the row header below the prompt (these areas
highlight when an attribute is held over them).
Students Students
BackpackWt
10.75 af ter 51
Group
S1 = mean before 79
Column Summary 130
S1 = count
By default, dropping a numeric attribute on one dimension gives you the mean; dropping a categorical
attribute gives you counts for each category; dropping numeric attributes on both dimensions gives you
correlation; and putting numeric attributes on one dimension and categorical on the other gives both
means and counts.
Students
BackpackWt
51
af ter
10.284314
Group
79
before
11.050633
Column Summary 130
10.75
S1 = count
S2 = mean
109
Fathom 2 Help
You can add a new formula (and show the formula editor) by choosing Summary | Add
Formula (see Add or Remove a Summary Table’s Formulas 111 ).
You can add a bunch of formulas at once by choosing Summary | Add Basic Statistics or
Add Five-Number Summary (see Add Basic Statistics 402 or Add Five-Number Summary 402
).
See also
Add or Remove a Summary Table’s Attributes 110
Use a Summary Table Without Attributes 112
Sort Categories in a Summary Table 112
Dropping on an
existing attribute
replaces that
attribute.
You cannot have both categorical attributes and numeric attributes along the same dimension (row or
column) of the table. Dropping a categorical attribute onto rows (or columns) when there are numeric
attributes present removes the numeric attributes, replacing them with the categorical attribute. Similarly,
dropping a numeric attribute onto rows (or columns) where there are categorical attributes present
replaces the categorical attributes with the numeric attribute.
To get correlation, drop a numeric attribute on one dimension when there are numeric
attributes on the other dimension.
To remove an attribute from a summary table, select it and choose Summary | Remove
Attribute.
110
How To’s
1. To add a formula to a summary table, select it and choose Summary | Add Formula.
2. Enter a function (or a more complicated formula) in the resulting formula editor and click OK.
You can add mean, count, standard deviation, standard error, and number of missing values all at once by
choosing Summary | Add Basic Statistics or add minimum, first quartile, median, third quartile, and
maximum by choosing Summary | Add Five-Number Summary.
Students Students
BackpackWt BackpackWt
10.75 2
130 8.5
4.2537034 10.5
0.37307446 13
0 24
S1 = mean S1 = min
S2 = count S2 = Q1
S3 = stdDev S3 = median
S4 = stdError S4 = Q3
S5 = count missing S5 = max
111
Fathom 2 Help
All of the people who were married at the time of the 2000 U.S. census are selected.
See also
Select Data in a Graph 37
Select Cases in a Case Table 72
no data
Drop an attribute here
31.006277
S1 =
If a summary table is connected to a collection, it will recognize attribute names in formulas you add to it.
To connect a collection to a summary table, drop its name on one of the summary table’s drop
spots.
Students
130
S1 = count
See also
Work with the Formula Editor 122
112
How To’s
Boston
Sex
Row
Male Female Summary
4 6 10
N/A
8100 4300 4300
75 85 160
ow nHom e Rents
23000 10276 16407.5
147 168 315
Ow ns or buying
23000 10400 16500
Column Summary 226 259 485
22700 10060 16000
S1 = count
S2 = median TotalIncome
See also
Sort Data in a Graph 61
Sort Data in a Case Table 74
A slider is a named value, or a parameter, that you can use in any formula in your document. A change
you make to the slider is immediately reflected by all the parts of your document that depend on that
slider. (see Sliders 339 .)
You might use a slider named proportion in a formula for an attribute named Votes so that changing the
slider value would change the proportion of randomly generated “yes” votes in the collection of voters.
Or you might use a slider named Coefficient in a function you plot on a graph. See Plot a Function You
Can Vary Dynamically (Using a Slider) 55 .
1. Create a slider by dragging one from the shelf or choosing Object | New | Slider.
By default, sliders are named V1, V2, V3, and so on. It’s best to give your sliders meaningful names that
reflect their role in your document. The name is selected when the slider is first created, so you can
113
Fathom 2 Help
change the names by typing the new name as soon as you create it .
2. Type a name for the slider. (Valid slider names follow the same rules as attribute names. See Tips for
Naming Attributes 330 .)
3. Use the slider’s name in the formula that you want to make dynamic. (If you don’t want to type the
slider’s name, you can find slider names in the formula editor’s list under Global Values.) See Work
with the Formula Editor 122 .
4. Change the slider’s value by dragging the thumb above the axis, clicking the slider’s animation button,
or editing the existing value by typing. See Animate a Slider 115 , Change a Slider’s Animation Speed 115
, and Restrict the Values a Slider Can Take 116 .
Although you usually first make and name a slider, if you have a formula editor open and you decide you
want to use a slider rather than a static value, you can type a reasonable name, and make the slider after
you’ve accepted the formula. At first you’ll get a #Name Not Recognized# error, but when you make and
rename the slider (assuming you give it the same name as you used in the formula), Fathom will recognize
it and use its value for the formula’s result.
To see examples of advanced slider functionality, look at Slider Examples.ftm
Subsections
Change a Slider’s Scale 114
Limit the Space Used by Sliders 116
Define a Slider by Formula 117
Create and Use Sliders Whose Values Have Units 117
Use a Slider to Compute Things Based on a Collection 118
Make a Slider Do a Random Walk 119
Make a Non-numeric Slider and Use It in a Filter 119
The numeric scale for a slider behaves just like a numeric axis of a graph. See Change Axes by Dragging
40 .
When the cursor is over the axis, you can drag to rescale or translate.
Or…
114
How To’s
Double-click the axis to open the slider’s inspector, and change the values for the appropriate
properties. See Slider Properties Panel 365 .
Or…
Edit the slider’s value; for example, giving a new slider (with default scale) a value of 1000
makes the bounds go from a bit below zero to a bit over 1000.
If the slider’s value is not defined by a formula, the slider thumb will move back and forth between the
ends of the slider axis, and the value corresponding to the slider will change accordingly. Any formula
containing the slider’s name will be recalculated, and all views of it will update.
To stop the animation, click the button. You can also stop all animations by pressing the Esc key on
your keyboard. (Sometimes you have to press Esc twice.)
While an animation is taking place, you can go ahead and do other things in Fathom (but the response
might be a bit slow).
See also
Change a Slider’s Animation Speed 115
Restrict the Values a Slider Can Take 116
115
Fathom 2 Help
See the sample document Slider Examples.ftm for advanced slider functionality.
See also
Slider Properties Panel 365
116
How To’s
1. With the slider selected, choose Edit | Edit Formula to show a formula editor.
2. Enter a formula. Here’s a fun one: randomInteger(1,6).
3. Close the formula editor and press the animation button on the slider.
The value of the slider will update each time you click the animation button. Notice that the slider’s
thumb has changed shape and is no longer draggable. Notice also that the value of the slider is gray to
remind you that it is computed.
You can edit a computed slider’s formula. Editing a slider value wipes out its formula; the text becomes
blue and the draggable thumb returns.
See the sample document Slider Examples.ftm for uses of the new slider functionality.
See also
Slider Properties Panel 365
Work with the Formula Editor 122
Slider values can have units, just the same way that attribute values can. See Attributes with Units 331 .
You can type a slider value with units, as shown above. Once the slider value has units, it will keep those
units while it animates.
A slider can also acquire units through its formula. The formula above will choose a random speed in
117
Fathom 2 Help
There are times when you want measures computed for a collection to be accessible throughout the
document. For example, suppose you have measured the weights of a bunch of butterfly cocoons, and
you want the mean and standard deviation of those weights to be used to generate simulated weights
from a normal distribution. You can use sliders to calculate the values you need and then refer to those
sliders in any formula in the document.
Here’s how:
1. Drag the name of the collection from the collection to the slider and release.
The slider will acquire a thick black border indicating that you can drop there.
2. Choose Edit | Edit Formula.
3. Enter a formula that refers to an attribute in the collection, such as mean(weight).
4. Edit the slider name to reflect its use.
Now you can refer to this slider in other situations. For example, in a dot plot of simulated cocoon
weights, you could plot values to compare the actual mean cocoon weight with the mean of the simulated
weights.
118
How To’s
Sliders are not required to have numeric values. You can give them string values, too. When you do, the
slider’s thumb disappears, and pressing the animation button doesn’t do anything.
Here is a bar chart of people from Ohio showing the metropolitan area from which they come. The slider
below the graph is used in a filter for the graph to limit cases to a single state. It is easy to change the
state by typing a new value for the slider.
See also
Write a Filter 121
119
Fathom 2 Help
There are times when you want to collect values from a slider along with other values from sensors or
keypresses.. For example, suppose you have a slider that generates speeds of automobiles as a random
function, and you want to collect those values in a collection. You can use sliders to generate the values
you need and then collect them in an experiment collection.
Here’s how:
1. Drag the plug from the slider to the collection and release.
The collection will change to an experiment collection and its' inspector will open to the experiment
panel.
2. Set up the collection parameters.(see Set Up an Experiment 154 )
See Also
Collect Values from a Meter 153
120
How To’s
4. Close the formula editor. The object will now show only cases where the filter is true.
Edit a filter by double-clicking it.
To remove the filter, choose Object | Remove Filter or edit the filter and erase it.
Filters are always Boolean expressions; see Make Boolean Expressions 130 .
For complicated filters, you may need Boolean operators (and, or, and not); see Operators 414 .
See also
Work with the Formula Editor 122
Make a Non-numeric Slider and Use It in a Filter 119
121
Fathom 2 Help
For example, if your student data has height expressed as two attributes, feet and inches, you might write a
formula for heightInInches, which would be 12 * feet + inches. If you had time and distance, you might make
speed.
You write the formula in the formula editor. A formula is not a complete equation; it is an expression for
computing the value of an attribute. We’ll use the example in which we define speed to be distance/time.
1. Make a new attribute to hold your calculated values (see Add Attributes in a Case Table 68 ).
2. In a case table for the collection, click the attribute name and choose Edit | Edit Formula. The
formula editor appears.
Alternatively, you can bring up the context menu (see Context (Right-Click) Menus 413 ) for the attribute
(as shown here) and choose Edit Formula. Or double-click in the formula box for the attribute. These
two methods work just as well in the collection inspector (see Cases Panel 354 ).
3. Enter the formula. Type just the right side; no equals signs. Use the asterisk (“*”) for multiplication
and the slash (“/”) for division. For our example, type distance/time. There’s a lot you can do with this
editor. For more, see Work with the Formula Editor 122 .
4. Click OK to close the formula editor. You should see the new values in the table column. (If you click
Apply , you can see the results in the table or inspector, but the formula editor stays open.)
Some attributes may require formulas with conditional functions such as if-statements. See Express a
Condition with an If-Statement 130 and Recode with a Switch Statement 131 .
Note: You can use the same context menu that gives you Edit Formula to use Cut, Copy, Paste, and
Clear with formulas. These commands are perfect for removing formulas from attributes (while
preserving their values) and for giving the same formula to many attributes.
Other uses of formulas:
Plot a Function on a Graph 53
Writing formulas for statistics (see Define a Measure 105 ) in a collection, or computing things in summary
tables, (see Add or Remove a Summary Table’s Formulas 111 ).
Writing filters. See Write a Filter 121 .
122
How To’s
Determine the values of a Define an attribute, PopulationDensity, and give it the formula Population/
new attribute. Area or one of the Random Functions.
Enter the formula sex = "M" to look at only the males in a collection of
Define a filter.
people. (see Write a Filter 121 .)
Suppose you wanted to throw dice until there were 2 sixes. You would use
Control a simulation. an “until” command in the Collect Measures panel of a measures
collection inspector with the formula count(face=6)=2.
To make a slider that will choose a random integer between 1 and 100
Compute a slider value. each time its animation button is pressed, give it the formula randomInteger
(1, 100).
With formulas, you don’t simply choose from a list of statistics or filters or simple computations. You
can compute practically anything you want, as long as you know how to express it.
See Fathom Operators, Functions, and Units 413 for a complete list of the functions available in Fathom.
Subsections
Open the Formula Editor 123
Color-Coding in Formulas 127
Use the Keypad 128
Use the Attribute and Function List 128
Make Exponents 129
Express a Condition with an If-Statement 130
Recode with a Switch Statement 131
Move the Cursor and Select a Portion of an Expression 129
Frequently Asked Questions About Formulas 135
123
Fathom 2 Help
Filter for an object (Restrict Select the object and choose Object | Add Filter.
the Cases You See by Filtering If there is already a filter showing underneath the object, double-click
Data 96 ) it.
Select a graph that has two numeric axes and choose Graph | Plot
Plotted function (Plot a Function.
Function on a Graph 53 ) or a
Most graphs that have at least one numeric axis can take a plotted
plotted value (Plot a Line at a
value. Select the graph and choose Graph | Plot Value.
Given Value on a Graph 49 )
For an existing plotted value or function, double-click its formula.
124
How To’s
A formula in Fathom (except for a filter) is simply the right side of your equation
or function. The left side is already entered for you in the formula prompt area.
What to enter
So if you want speed = distance/time, open the formula for speed and enter distance/
time, as shown.
You can enter your formula on the keyboard if you wish; what you type goes
The keyboard
right into the formula pane. See Keyboard Shortcuts 441 for details.
The title bar describes what you are writing a formula for; in this case, the
Title bar
attribute speed. Drag the bar to move the editor around in the document.
You enter your formula here. You can enter it by typing, clicking on the keypad,
Formula pane
and/or double-clicking items in the attribute and function list.
Formula prompt This is where Fathom writes the left side of the equation.
Choose from this pop-up menu to make the formula font bigger or smaller.
Font size pop-up
Your choice will be remembered as a preference.
Adjust this bar up or down to give yourself less or more room in the formula
Resize formula bar
pane.
Click these buttons to enter numbers and commonly used operators. Hold down
Keypad
the Ctrl (Win) Option (Mac) key to see some additional keypad choices.
125
Fathom 2 Help
This pane lists the attributes and measures available to you as well as the many
functions Fathom recognizes. Expand the lists to see the choices. To enter a
function or attribute, double-click it. To learn what a function does, select it
Function and (single-click) and look in the help pane below.
attribute pane Here is information on Fathom’s built-in functions: Fathom Operators,
Functions, and Units 413 .
Here is information on Fathom’s Special Values 434 .
The Cancel button closes the formula editor without applying any changes since
the editor was opened or since the Apply button was pressed. Apply makes the
Buttons
current expression in the formula pane take effect. The OK button applies the
formula and closes the formula editor window.
Resize help bar Drag this up or down to make the help pane bigger or smaller.
Displays help text for currently selected heading or function in the function and
Help pane
attribute pane.
Resize the formula editor to suit your need to have a larger or smaller formula
pane, function and attribute pane, or help pane. On Windows, you can resize by
Resize window
dragging the edges or corners of the window. On Mac, drag the lower-right
corner of the window.
To get … Type …
pi
width^2
x+1, then
,
, then /2; alternatively, use parentheses: (x+1)/2
x^2, then
, then +1
Use * for multiply, / for divide, and ^ for exponentiation.
Most familiar functions work using the most common abbreviation: sin, random, floor, ln, log,
126
How To’s
asin, exp, sgn, to name a few. Functions require parentheses around their arguments, as in sin
(x).
Use to get the cursor out of an exponent, root, or the denominator of a fraction.
Did π not appear when it should have? When you type pi and then immediately type another
letter or a number, Fathom doesn’t know that you mean pi to be its own word. After typing pi,
press one of the arrow keys or the space bar (or * if you want to multiply). After the π appears,
you can type anything you want, and the π will remain π.
Entering one vertical bar “|” makes a pair of absolute-value bars.
The double-quote key makes a pair of quotes.
The open parenthesis “(” makes a pair of parentheses.
When you type if( you get a complete (though empty) if-statement, with three parts for you to
fill in. Type the condition (for example, income > 100000), then press Tab to move to the
results (result-if-true goes on the top, result-if-false on the bottom). If you want the results to
be words (also known as strings), you must surround them with quotation marks.
When you have more than two results, instead of nesting if-statements, you can use the switch
function. This is useful for recoding attributes. It takes an optional expression inside
parentheses and then evaluates each of any number of true/false expressions to determine
which value to return. To add an alternative, press Insert (Win) or Option+Return (Mac) on
your keyboard.
To insert something in the beginning, don’t click to place your cursor; instead, arrow to the
beginning before adding to the existing formula.
127
Fathom 2 Help
Constants Red
(true and false are constants.)
Units Green
(m and s are the units meters and seconds.)
To see a list of choices in a given category, click the open/close control next to the category. In this
example, the user is about to open the Attributes list.
To enter an item from the list into the formula, double-click it.
To close a list, click the open/close control.
To see an explanation of an item, click it once to select it. You can then read the help for that
item in the help pane at the bottom of the editor. This is especially handy if you can’t
remember the exact syntax for using a function.
When you want to see more of various portions of the formula editor, for example, when you have a tall
formula or you want to see more of the formula list, you can drag the borders of the formula and help
panes. (You cannot make the middle portion too small to see the keypad.)
128
How To’s
Selection can play an important role in creating formulas. When an expression is selected, pressing certain
keys causes them to operate on the entire selection. For example, if a + b, without parentheses, is selected
and you press * (for multiplication), you get (a + b)*.
In each of the following examples, the user types /4.
129
Fathom 2 Help
We’re talking about “true” and “false” here and how to combine them. For example, if you’re looking at
census data and you want to find women over 60 who are either married or divorced, then this expression
will filter out all the other people in the census file:
The comparison operators =, <, and > can be found on the formula editor’s keypad. You can also find
≤, ≥, ≠ by holding down Ctrl (Win) Option (Mac).
Express the not of an expression (the negative of an expression) by positioning the cursor in
front of the expression and clicking the not key on the formula editor keypad.
You can use and and or to string logical expressions together.
When in doubt about which expressions are evaluated first, use parentheses to force the
evaluation order you want.
“True” and “false” are allowed values for attributes, so you can write a formula for an attribute
that returns a Boolean value.
From the keyboard:
For and, make sure the preceding expression has parentheses around it and type the word
"and".
For and, type &. This has the same effect as the keypad button.
For or, make sure the preceding expression has parentheses around it and type the word or.
For or, type Ctrl+Shift+O (Win) Option+Shift+O (Mac) (that’s the letter “O” as in “Oscar”).
For not, type ~. (~ is the tilde key, usually in the top left of your keyboard.)
To create an if-statement,
Type if( Fathom fills in the other parenthesis, the brace, and a question mark for each of the
values to be filled in;
130
How To’s
Or …
Open the Function list to the Conditional category and double-click if.
Enter the condition then use Tab to move to the next field (the “if true” result on top, “if
false” on the bottom).
You can create nested if-statements by using if-statements as values for if-statements. Usually, this is
more efficiently done using the switch function (see Recode with a Switch Statement 131 ).
131
Fathom 2 Help
Usually you will write switch statements similar to the preceding one. Your value to switch on will be the
name of an attribute, the test values will be values that the attribute takes on, and the result values will
amount to a recoding of the original attribute. Fathom returns the result value whose test value matches
the value being switched on.
A switch statement’s tests are not limited to equality. In this statement, for example, Fathom first
compares the value of N with 0 and returns “zero” if N does equal 0. If not, it tests to see whether N is
less than 5 and returns “low” if it is. If not, it goes on to compare N with 10 and returns “medium” if N
is less than 10. If no test has succeeded thus far, it returns “high.”
Here are some things to keep in mind when writing switch statements.
If you do not explicitly specify a comparison operator, Fathom will test for equality.
Fathom returns the first result value whose test returns true. The fact that later tests might also
return true has no effect on the result.
A switch value can itself be an expression, not just the name of an attribute, for example,
circum
diameter . The result of evaluating the expression is used in the test.
Result values may be any expression (including if- or switch statements).
The switch value is substituted for any question marks or missing arguments to functions that
appear in the test expression.
Much data come with categories coded numerically rather than as words. But when you’re doing data
analysis with Fathom, you want the categories to be words you can read, not numbers you have to keep
translating.
132
How To’s
Recoding is done with the switch function, which is a conditional function that acts like a complicated
if-statement.
The collection here has two numerically coded attributes: sex: 1 = "Female", 2 = "Male"; education: 1 =
"High School", 2 = "College No Degree", 3 = "College Degree", 4 = "Graduate Work"
1. Add two attributes: sexDecode and edu.
2. Double-click the formula cell for sexDecode to show the editor.
3. Enter this formula:
Tab to move in the switch function. To insert the second option, press Insert, or Ctrl+Enter (Win)
Option+Enter (Mac).
You should include the “Other” value in the formula so that wrongly entered data and missing values
don’t get coded as male.
4. Similarly, give edu this formula:
People
s e x e ducation s e xDe code e du <ne w >
1 1 3 female college degree
2 2 2 male some college
3 1 1 female high school
4 1 4 female graduate w ork
5 1 4 female graduate w ork
6 2 3 male college degree
7 1 3 female college degree
8 2 4 male graduate w ork
9 2 1 male high school
10 2 3 male college degree
11 1 4 female graduate w ork
12 1 2 female some college
13 2 2 male some college
You could now clear the recoding formulas and even delete the original attributes.
133
Fathom 2 Help
By default, Fathom will use the units that appear in the original
attributes in the result. But you can change this by typing a new,
compatible unit in the units row. For example, you could change m/s
to mph.
134
How To’s
One thing to be careful about is naming attributes with unit names. For example, if you name an
attribute seconds, Fathom is going to prefer the attribute over the unit in formulas. This can be confusing!
So, the best idea is to keep units to the units row and out of the attribute name.
See also
Attributes with Units 331
Manage Units in a Case Table 77
Create and Use Sliders Whose Values Have Units 117
Manage Units in the Collection’s Inspector 92
I see things I want to use in the list. How do Double-click something from the list to get it into the
I get them into the formula? formula.
I’m having trouble editing a number. I can’t With the cursor on either the left or right side of the
click in the middle of it. What do I do? number, use the arrow keys to move inside the number.
Pressing the comma doesn’t work to get to You don’t have a correct function. Make sure the
the second parameter of a function. Why function name is colored blue, which means Fathom
not? recognizes it.
Do I have to be careful about uppercase and Except for units, no. For example, stddev works as well
lowercase? as stdDev.
Are string comparisons case sensitive? No. "New York"="new york" returns true.
Why is it when I type mhz for megaHertz, Units are case sensitive. The abbreviation for megaHertz
Fathom doesn’t recognize it? is MHz with a capital M and capital H.
135
Fathom 2 Help
I’m not sure about Fathom’s order of Fathom obeys the normal rules of algebra. But, when in
evaluation. How does it work? doubt, use parentheses.
With Fathom’s Survey extension, you can use Fathom to create and upload a data-gathering form to the
web, and download the results into a Fathom collection for analysis. Here are some examples of
classroom use.
In a science class, each group enters the
data for each pendulum measurement it
makes.
136
How To’s
Subsections
Limitations and Cautions 137
Fathom Surveys drastically streamlines the process of gathering, analyzing, and disseminating
information. When it is used to gather personal information, there is real potential for harm. (see the
Ethics of Classroom Surveys online tutorial that comes with Fathom Surveys.) The Fathom Surveys
License contains additional cautionary information.
137
Fathom 2 Help
1. Define the attributes that will hold the data you want to collect (each attribute will correspond to one
question on the survey).
2. Type the wording of each question in that attribute’s Question field. If you don’t type a question, the
attribute name will appear in the survey.
Tips
Resize the inspector by dragging its edge. By default, questions appear on one line in the
inspector; to resize a row drag the boundary below the attribute name.
By default, survey questions have free-response answers. To give respondents specific answer
choices, see Add Multiple-Choice Answers 139 .
1. Type any instructions or introduction to your survey in the Instructions pane. When the survey is
complete, click Upload Survey. (You’ll be asked for your username and password. See Upload a
138
How To’s
By default, respondents are provided with a text field into which they can type their responses. It’s often
desirable to instead give them a pop-up menu of choices (called a category set in Fathom). Here’s how:
Click in the Format field for the attribute and choose Define New Category Set from the pop-up menu.
Type the options in the order you want them to appear, with commas between items.
139
Fathom 2 Help
3. Click OK.
Once you have defined a category set, it appears in the Format pop-up menu, allowing you to apply it to
more than one attribute. (For example, you can have a category set with the options no response, I
love it, I like it, I don’t care, I don’t like it, I hate it, which you can use for several questions.)
See also
Click the View Survey link in the inspector to open your web browser to the page with the
survey form. Right-click (Win) Ctrl-click (Mac) the link to copy it.
To make changes to a survey form (such as adding instructions or pop-up menu answers),
1. Make changes to the Fathom document you used to create the survey (see Create a Survey 138 , Add
Multiple-Choice Answers 139 for more detail).
140
How To’s
Among the things you can do are: add or delete questions, change the wording of questions, change
the title of the survey, and change the answer format for one or more of the questions.
2. In the collection’s Survey panel, click Upload Survey. (If you have closed the document since you last
uploaded it, you will need to give your username and password again.)
3. Remember to save your changes.
See also
A newly created survey cannot be seen or taken by students until it is published by the instructor. This
gives the instructor necessary control over access to a survey. For example, a teacher might not want
students in a class to see a survey until class begins. Or, a student who can upload won’t be able to have
respondents begin taking a survey until the teacher, as instructor, has approved and published it.
Instructors can determine in their preferences whether surveys they create are immediately published by
default or left unpublished (see Update User Information 150 to learn how to change this preference).
Surveys created by students always must be approved by instructors before they are published (see
Approve Surveys Created by a Student 143 for more information).
To publish a survey:
1. Go to the Fathom Surveys website.
2. Navigate to the list of surveys and click the Edit Properties icon for the survey you wish to publish.
3. On the Properties page click the check box next to Publish Survey.
141
Fathom 2 Help
Do you want students to be able to see the results of a survey? If so, you need to publish those results.
But if the responses to a survey are not meant to be shared, as would be true of a pop quiz or a student
project, then the instructor must make sure the results are not published. Instructors can determine in
their preferences whether surveys they create have their results published or not (see Update User
Information 150 to learn how to change this preference).
To publish the results of a survey:
1. Go to the Fathom Surveys website.
2. Navigate to the list of surveys and click the Edit Properties icon for the survey you wish to publish.
3. On the Properties page click the check box next to Publish Results. Click the Apply Changes button.
Students will now be able to see the results of this survey and download these results to Fathom.
See also
Suppose you have a collection of data in Fathom, and you want respondents to add to that data using a
web form.
1. Select the collection and choose Collection | Create Survey. Go
to the Survey panel of the collection’s inspector and check Replace
Data On Site
If you want a customized survey form, you will need to host the survey on your own (or school’s)
website. (The survey results will still be kept on the Fathom Surveys website.) You will be editing the
HTML code, so you need a program that will let you do that. Unlike surveys hosted on the Fathom
142
How To’s
Surveys website, after submitting your form, respondents will not see an empty form ready for another
submission, but will see either the successful submission page or the unsuccessful submission page. You
will need to create those pages.
1. Create the survey in Fathom and upload it to the Fathom Surveys website.
2. Log in to your account. Locate the survey and click the html source button.
This gives you a window with a text box containing the HTML for your survey. All of the text is selected.
3. Copy the contents of the text box, and paste it into a text or HTML editor.
4. Locate this text: “hidden name” = “success forward” value = “ ”. Between that last pair of quotes, type the
URL for the successful submission page.
5. Similarly, locate this text: “error forward” value = “ ”. Between that pair of empty quotes, type the URL for
the unsuccessful submission page.
6. Edit the code for the page to create the look you want.
Warning! Do not edit the name and password fields or any of the name = question number fields, or
the data collection won’t work.
7. Upload the edited HTML to your own website. Note on Microsoft’s FrontPage: Use of FrontPage for
editing your survey is not recommended. If you must do so, paste the HTML into FrontPage using
the Paste Special: Paste as HTML command. All of the built-in formatting will be gone (for example,
questions and answer fields are run together, rather than each question having its own line).
Give the URL for the Take Survey page to respondents. Make sure they know their username and
password. You can copy the URL from your browser or you can right-click (Win) Ctrl-click (Mac) the
View Survey link on the Survey panel in the Fathom document.
Respondents following the URL will be asked to log in before they are sent the Take Survey page.
Different people can all be filling out and submitting data from different computers at the same time; it
will all become part of the same set of results. If students are sharing computers or entering multiple
experimental results, they can simply fill out the form over and over, clicking Submit for each person or
set of results.
See also
View and Download the Results 144
143
Fathom 2 Help
instructions on how to do this). The purpose of this step is to make sure that a responsible adult reviews
the survey before it is made available online.
If you have the survey-creation file open, click Download Results in the Survey panel of the collection’s
inspector. The collection fills with all data collected on the website. You can make graphs, summary
tables, or other objects from this data, just as you would with any Fathom collection.
If you don’t have the survey-creation file, go to the list of surveys on the Fathom Surveys website, click
the View Results link, and then drag the URL and drop it into an empty Fathom document.
Suppose the data you get from a survey is messy and needs to be cleaned up before it can be analyzed.
You can clean up the data in the survey file’s collection, and replace what is stored on the Fathom
Surveys website with the cleaned up version of the data.
144
How To’s
You can leave a survey intact, but remove the results (to re-use a survey form with another group, for
example).
1. In Fathom, select the survey collection and choose Edit | Select All Cases.
2. Choose Edit | Cut Cases. (You can then paste them into another collection, if you want.)
3. In the Survey panel of the collection’s inspector, check Replace Data On Site and upload the survey.
The survey form will remain intact, but the results will be cleared.
If you want to get rid of the survey, see Delete a Survey 145 .
1. Click in a blank place in the document or make a new Fathom document if you want the surveys in
separate documents.
2. Choose Edit | Paste Collection. Rename the collection by double-clicking its name and typing a new
name.
3. If the original collection had cases in it, you should delete them in the new collection by choosing Edit |
Select All Cases, then choosing Edit | Delete Cases.
The new collection retains all the survey information of the old collection, but it has not yet been stored
on the Fathom Surveys website. You can change it however you wish.
1. Log in to the Fathom Surveys website (using an instructor username and password).
2. Locate the unwanted survey and click the delete survey icon.
If you want to keep the survey but get rid of the results, see Delete Survey Results 145 .
145
Fathom 2 Help
You can sort surveys in your list on the Fathom Surveys website. To do so, click on a column name. Click
on the column name a second time to sort by that column in the opposite order. For example, clicking
twice on Survey Published will sort the surveys so that all the published surveys are at the top.
This help system for Fathom Surveys will be installed, including the movies.
You will be sure that users will be running Fathom version 2.03 or later, which means that when
students download survey results, they will get a Survey panel that includes the questions that
belong to the survey.
146
How To’s
Subsections
Explanation of User Roles 147
Log in to Your Fathom Surveys Account 148
Add Instructor Users 148
Add Student Users 148
Allow a Student to Upload Surveys 149
Update User Information 150
Change Administrators 151
Administrator Teacher, Authorize new instructors; Upload surveys, Receives an email from
administrator see all instructors. add students, Key Online with
, or computer publish surveys information needed to
technician or survey results get started. Initiates
authorization of new
Teachers using Key
Online.
Instructor Teacher or Upload surveys, fill out Add/delete Receives email from
(less often) surveys, publish surveys and instructors, view Administrator with
administrator survey results, delete or modify information needed to
surveys, add/delete anything get started.
students, update student belonging to
information and own user other instructors
information
Student Student or Fill out published surveys, Publish survey Typically one student
Anyone see and download published or survey account is used by many
without survey results. May be results, modify students to take
Fathom granted ability to upload anything on the surveys.
access surveys from Fathom. website.
147
Fathom 2 Help
To access your instructor account, go to Key Online and click the My Account tab or follow the View
Survey link in an uploaded survey.
To create student accounts (for example, one student user for each class you teach), you must have an
instructor account.
148
How To’s
Keep track of passwords you assign to students. You cannot retrieve a password from Fathom Surveys if
you forget it, but you can always assign a new password.
See also
149
Fathom 2 Help
Follow the Students link after logging in. Click on the edit
properties icon for the desired user. Change the information for the
given user and click Apply Changes. (Note that if you don’t wish to
change the user’s password, you should leave the two password fields
empty.)
150
How To’s
There are many different ways to use experiments in the classroom. Here are some examples:
Light: Capture a waveform. Use a light sensor to analyze the differences in intensities in light from an
LCD monitor, fluorescent lighting, a candle, and sunlight.
Temperature: Compare the cooling rates of different liquids such as oil, water, and soda. Compare
changes in temperature relative to changes in density. Compare the heating and cooling curves for
different liquids.
Velocity and Acceleration: Roll different objects through photogates on ramps. Compare differences in
acceleration relative to mass, height of ramp, inclination of ramp, and distance rolled.
Events and Timers: Record the number of cars passing a house next to a school. Compare the
distributions of arrivals just before and after school and during the day.
Subsections
Work with Meters 151
Connect Sensors 153
Set Up an Experiment 154
Subsections
151
Fathom 2 Help
See also
Set Up an Experiment 154
If there are no sensors plugged in, click on the meter icon in the shelf. A pop-up menu appears. Select
a meter from the list and drop it into your document.
If there is a sensor plugged in, click on the meter icon in the shelf. An object will appear for the
connected sensor; drop it into your document.
Go to the Object | New menu and choose the type of meter you want. If a sensor for this meter is
plugged into the computer, the meter name will appear above the line (see Connect Sensors) 153 . Drop
it into your document.
If Fathom is running when you plug in a sensor, a new meter for it will be automatically created in the
active document. If there is already a meter of the correct type in the document, then the sensor will be
assigned to that meter instead (see Connect Sensors) 153 .
Connecting Photogates
If you have a photogate connected but the meter name does not appear above the line or you do not get
the automatic meter appearance, then your photogate is a non-auto-id photogate. In this case, you need
to select the photogate meter from the list under the line in the menu. Fathom also assumes that the first
photogate meter corresponds to a photogate in Channel 1. If you have only one non-auto-id photogate
152
How To’s
Go! Devices: The sensor cable plugs directly into a USB port on
the computer.
Go! Link: The sensor plugs into a Go! Link interface that plugs
into a USB port on the computer.
153
Fathom 2 Help
Connect Photogates
If you are connecting only one photogate then make sure that it is attached to Channel 1. There is no way
to specify or change the channel that a sensor is associated with. For all auto-id sensors, the channel is
automatically detected and moving the sensor will reassign the channel. Photogates are largely non-auto-
id sensors and do not communicate channel information to the computer.
Experiments can also be set up without the sensors physically present; for example, by the instructor at
home to facilitate data collection in the classroom, or as homework in preparation for a lab period.
To collect data at regular timed intervals, click Collect By Time. You can specify how many cases you
want to collect, at what rate, and for what duration. If you specify any two of these then Fathom
calculates the third quantity, for example, 100 cases per second for 3000 cases, or 100 cases per second
for 30 seconds. If you collect by timed intervals, time is automatically an attribute of the experiment.
If you are trying to collect data at a very fast rate (faster than 100 cases/second) from a LabPro sensor,
for example, 10,000 cases/second from a microphone, then data collection proceeds in batch mode. This
means that no cases will be visible in the experiment collection until the experiment is over. The meters
will update during this duration and the experiment progress bar will indicate that the experiment is in
progress. The LabPro buffers the data in batch mode, so there is a limit on the total number of cases
154
How To’s
collected. The LabPro buffer can only store 12,287 points, and this is divided by the number of attached
sensors to give the limit on number of cases. For instance, if you have three sensors attached, the LabPro
will store about 4000 cases.
If you are collecting faster than 10,000 cases/second (especially with triggering) then collection proceeds
in a Fast Mode. This mode may have some channel limitations. If you experience problems, make sure
your sensor is attached to channel 1 of the LabPro.
To collect data based on keyboard events, click Collect By Keypress. If you collect by keypress,
Fathom always collects the value of the key pressed.
To collect data from a photogate, create a photogate meter and attach it to the collection. An
experiment inspector will appear. A photogate does not have to be present to create the experiment.
(see Collect from Photogates 153 ).
Click Turn Experiment On to begin collecting. You can defer the start of the collecting based on
some external condition (see Triggering 155 below).
155
Fathom 2 Help
3. Double-click in the Formula column if you want to define the attribute by formula.
See also
Add Attributes in a Case Table 68
There are many ways to build simulations in Fathom, but they all follow this basic pattern. Below is a
simple coin-flipping simulation in which the results are the number of heads when three coins are flipped.
Subsections
Sampling Simulations 157
Collect Measures and Other Results 159
Use Scrambling to Test for Independence 162
Simulation Examples 163
156
How To’s
See also
Random Functions 421
Collect Measures 383
Regular Sampling
One reason to sample is that you have a big collection and you only want to look at a little of it. So you
choose some cases at random and use them as a stand-in for the whole population.
Simulation
Another reason to sample, especially while you’re learning, is to do simulation. You can learn about the
mathematical properties of sampling by repeatedly sampling from a population to see what the samples
would have told you if you had seen each of them alone. You might use this, for example, to study
confidence intervals.
You can also use sampling to simulate common (and less-common) probability events. For example, to
make a coin-flipper, make a collection with two cases, heads and tails. Then sample from it with
replacement.
Bootstrap
Suppose you have N cases in your collection.
If you set up the sample collection to make a new collection by drawing N cases from the original
collection with replacement, this is called a bootstrap. This technique draws as many items as were in the
original collection, but because it’s with replacement, the exact distribution may not be the same. Some
cases will have been chosen twice; others not at all.
You often use a bootstrap to establish a confidence interval for a summary statistic (the median, say).
See also
Take a Sample 157
Collect Measures and Other Results 159
Generate a Sampling Distribution 164
4.9.1.1 Take a Sample
Sampling in Fathom means to choose some number of cases randomly from a collection and put them
into a new collection. The original collection is not changed. The sample collection is a kind of derived
collection. There are two ways to sample cases from a collection: Drag and drop or use a menu
command.
2. Drag the name of the source collection into the new collection.
The new collection becomes a sample collection.
157
Fathom 2 Help
Using a Menu
1. Select the collection.
2. Choose Collection | Sample Cases.
A sample collection is created and filled.
You can control the sampling process in the Sample panel of its inspector (see Sample Panel).
If you Define a Measure, you can collect measures to build a sampling distribution of your measure(s)
(see Collect Measures and Other Results 159 ).
See also
Change the Number of Cases Sampled 158
Collect Another Sample 158
158
How To’s
Resize the sample collection so that you can see its cases. You will also see a button in the top
right corner labeled Sample More Cases. Click this button to collect a new sample.
With the sample collection selected, press Ctrl+Y (Win) z+Y (Mac), which is the shortcut for
collecting another sample.
Select the sample collection, and choose Collection | Sample More Cases.
Open the sample collection’s inspector, go to the Sample panel, and click the Sample More
Cases button.
See also
Sample Panel 359
159
Fathom 2 Help
A measures collection is a derived collection (see Derived Collections 340 ) whose attributes correspond to
measures in its source and whose cases correspond to individual runs of the simulation. To make a
measures collection:
Drag-and-Drop Method
1. Make a new, empty collection, for example, by dragging one off the shelf.
2. Drop a measure’s name onto this new collection. (Drag it from the source collection inspector’s
Measures panel.)
Menu Method
1. Select the collection from which you want to collect measures.
2. Choose Collection | Collect Measures.
The measures collection appears, and five sets of measures (the default) are collected. The name of the
measures collection tells you what it is derived from. You can use a measures collection in all the ways
you would any other collection: make case tables, graphs, and summary tables from it; assign new
attributes computed from the collected measures; and so on.
Control this process in the Collect Measures panel of the measures collection’s inspector
(see Collect Measures Panel 360 ).
See also
Collect More Measures 160
Simulation Examples 163
160
How To’s
When a statistical object, such as a test, is analyzing a collection whose values are generated randomly, it
becomes useful to collect the results from the test for repeated rerandomizations. This allows you to
build distributions from such things as linear-regression slopes, t-statistics, p-values, and so on. This dot
plot shows the results of collecting 100 p-values under conditions in which the null hypothesis is true.
You can also use sliders in statistical objects and then collect measures for different slider values.
1. Select the statistical object. (It should be one whose results are going to change either manually or
automatically.)
2. Choose Test | Collect Results as Measures.
You will get a new collection with five cases (the default number) whose attributes correspond to
161
Fathom 2 Help
1. Select the collection whose attribute values you would like to scramble. (It doesn’t make much sense
to do this unless there are at least two attributes in the collection. Usually, there is a measure whose
sampling distribution you are interested in.)
162
How To’s
2. Choose Collection | Scramble Attribute Values. By comparing the case tables for the collections,
it’s clear that the first of the attributes has had its values scrambled. The inspector for the scrambled
collection has a Scramble panel (see Scramble Panel 363 ) with which you can choose which attribute
is the one that is scrambled.
See also
Test for a Difference in Medians Between Two Groups 170
A deck of cards is made of spades only. The cards are shuffled, and the first card is drawn. The value is
written down, the card put back in the deck, and the deck shuffled again. A second card is drawn. What is
the probability that the two cards were the same? Theoretically, the probability is 1/13, because it doesn’t
matter what card is drawn first and the probability is 1/13 that the second card will match the first. Let’s
see how to construct a simulation of the situation.
We’ll start with a collection of just two cases, one for each drawn card. (A sampling simulation would be
more faithful to the scenario but slightly more complicated to construct.)
1. Make a collection with two cases and an attribute cardValue with the formula randomInteger(1,13)
Now we need a measure that detects when the two cards are equal. We’ll use this opportunity to
introduce the extremely useful uniqueValues function.
2. In the Measures panel of the collection’s inspector, make a measure named IsAPair and give it the
formula uniqueValues(cardValue)=1
163
Fathom 2 Help
Convince yourself that this measure does the right thing by rerandomizing a bunch of times (Collection |
Rerandomize). You could, of course, record the results of rerandomization by hand, but the beauty of
using a computer is that you can automate the process. In Fathom, that’s done by collecting measures.
3. Make an empty collection and drag the IsAPair measure from the inspector onto the empty collection.
You should see five gold balls fly from the collection to the newly named Measures from Two Cards
collection. This measures collection now has five cases with a single attribute named IsAPair whose values
are “true” or “false.”
4. Make a bar chart of IsAPair from the measures collection.
We want Fathom to collect many more measures so that we can get a reasonable estimate of the
probability of a pair.
5. Show the measures collection’s inspector, and go to the Collect Measures panel.
6. Change the 5 to 95 for the number of measures, and click Collect More Measures.
As the measures are collected, you’ll see the bar chart respond. How does the observed probability in this
simulation compare with the theoretical value?
See also
Generate a Sampling Distribution 164
Make a Coin-Flip Simulation 166
Test for a Difference in Medians Between Two Groups 170
164
How To’s
precisely, the histogram approximates the sampling distribution, which is the distribution you would
obtain from all possible samples.)
Some statistics, especially those commonly used in statistical practice, are known to have sampling
distributions that can be approximated, under certain conditions, by a normal or other probability
distribution. For many statistics, however, the only way to come up with the sampling distribution is to
use a computer to actually carry out the repeated sampling procedure.
In this example, we create the sampling distribution of the median age of people from a certain collection
of individuals from the 2000 census. We consider the collection of 500 ages to be the population. The
distribution of ages in the population is shown here. We’ll draw simple random samples of 10 ages and
compute the median age for each sample.
1. Start with a collection. (You don’t have to start with census microdata, but if you want to, the
instructions are here Import U.S. Census Microdata from IPUMS 30 .)
2. With the population collection selected, choose Collection | Sample Cases.
This creates a new collection with 10 cases chosen randomly with replacement. If you want to change the
number of cases per sample, change the sampling controls in the Sample panel of the sample collection’s
inspector.
3. Double-click the sample collection to show its inspector, then choose the Measures tab.
4. Define a new measure, as shown, that computes the median age of the people in the sample. (You can
define as many measures as you want.)
165
Fathom 2 Help
The section Have Fathom Randomly Generate Data 27 showed how to flip a coin using a function for
choosing at random. Here we’ll flip coins by sampling.
Suppose we are interested in the question “On the average, how many times will we have to flip a coin
before we get three heads in a row?”
1. Make a collection with one attribute, Face, and two cases, “H” and “T.”
2. Name this collection Coin. It represents the coin and its two faces.
166
How To’s
4. Show the inspector for the sample collection, and go to the Sample panel.
5. Configure the Sample panel as shown. (Double-click the formula area to create an Until formula.)
The Until condition tells Fathom to keep sampling until the three most recent values for the attribute Face
are all “H.”
6. Go to the Measures panel.
7. Create a new measure named NumberOfFlips.
8. Double-click the Formula cell for NumberOfFlips, and give it the formula count().
The formula count(), when you don’t specify count of what, gives the number of cases in the collection, in
this case, the number of flips it took to get three heads in a row.
9. With the sample collection selected, choose Collection | Collect Measures.
You should see a new collection named Measures from Sample of Coin. Fathom collects five measures
immediately.
167
Fathom 2 Help
Because we didn’t give the prev function an initial value, the molecule will start at point (0, 0). It hasn’t
gone anywhere yet; there are no steps or cases.
4. With the case table selected, choose Collection | New Cases.
5. Specify that you want 100 new cases and click OK.
168
How To’s
6. Name the collection Molecule, because its cases represent the position of the molecule over 100 time
intervals.
Your case table should look similar to the one shown above.
7. Make a line scatter plot, as shown.
This graph represents the motion of the molecule. Each time you choose Rerandomize from the
Collection menu, you see a new possible path. If you adjust the axes to go roughly from -10 to 10 in
both dimensions, you’ll usually be able to see the whole “walk” after each rerandomization (if you leave
the axes at their default, you have to keep rescaling the axes to see each random walk).
8. Show the inspector for the Molecule collection.
9. In the Measures panel, define a new measure called distance.
We want this measure to compute how far the molecule is from the origin in the last case.
10. Double-click the Formula cell for the distance measure, and give it the following formula:
By squaring each final position, then taking the square root of their sum, we get the distance from the
origin, regardless of the direction.
11. With the Molecule collection selected, choose Collection | Collect Measures.
The molecule collection rerandomizes five times and a new collection, Measures from Molecule, appears.
It contains the final distance from the origin for each walk.
12. Double-click it to show its inspector.
13. Set the inspector’s Collect Measures panel to collect 100 measures, and click Collect More Measures
.
Here is a dot plot showing 100 sets of 100 steps each. How do you think the result would differ if,
instead of 100 steps, we allowed 1000?
169
Fathom 2 Help
Here we work through a simulation in which scrambling helps us decide whether or not an observed
relationship between two attributes is significant.
Suppose we have conducted an experiment in which we drop crumpled-up paper from a certain height
above a target. After each drop, we measure the distance of the paper from the target. There are two
kinds of paper, copier paper and paper towels, and we are interested in whether there is a difference in
the distances they land from the target.
The data are shown here. To work through this example, create this collection (or use a similar collection
of your own data).
We’ll use the difference between the two medians, 3.6 in., as our measure of difference. How likely is this
value to occur by chance?
1. Define a measure in the OrbEx collection as shown below.
170
How To’s
2. With the OrbEx collection selected, choose Collection | Scramble Attribute Values.
A new collection, Scrambled OrbEx, is created.
3. Graph distance split by paper for the scrambled collection.
In the scrambled collection’s inspector, notice that the diffMedians measure is defined and that it computes
the difference of medians for the two groups.
4. Look at the Scramble panel in the scrambled collection’s inspector. See Scramble Panel 363
You could have used the pop-up menu to scramble distance instead of paper. (Would it make a difference?)
5. With the scrambled collection selected, choose Collection | Collect Measures.
The results of 100 scramblings are shown. Of the 100 scramblings, 11 have a difference of medians
greater than those in the original collection; thus we cannot rule out chance with much confidence.
Fathom has three statistical inference objects: interval estimate, hypothesis test, and linear model.
1. To use a statistical object, drag one from the shelf or choose Object | New [object].
2. Choose the specific type of object from the pop-up menu in the object’s upper-right corner.
All options for these objects accept attribute drops from a collection.
All types of estimate, as well as most tests, can also accept summary data by allowing you to edit the
default information (the bold, blue text fields).
Here’s the list of inferential procedures Fathom includes:
171
Fathom 2 Help
Interval Estimate: Estimate Mean, Difference of Means, Estimate Proportion, and Difference of
Proportions
Hypothesis Test: Test Mean, Compare Means, Analysis of Variance, Test Proportion, Compare
Proportions, Goodness of Fit, Test for Independence, Test Slope, and Test Correlation
Linear Model: Simple Regression and Multiple Regression
In addition to inputting summary data, you can link fields of these objects to sliders, to see how changes
in such things as confidence level or standard deviation affect results. Furthermore, you can collect
measures from these objects and graph the relationships.
How do you remember which object you want? If you want to compute a confidence interval for
something, that’s an estimate. If you want to test the statistical significance of a hypothesis, that’s a test. If
you want to do simple or multiple linear regression, use a model.
For more detail on working with these objects, see the appropriate topic within this section.
See also
Use Raw Data in Statistical Inference Objects 172
Enter Summary Statistics in Statistical Inference Objects 172
Estimate Parameters and Compute Confidence Intervals 174
Test Hypotheses 183
Build a Linear Model 202
172
How To’s
You can use statistical inference objects to explore what-if’s, such as, “What happens to the p-value when
the standard deviation changes?” or “How does sample size affect the width of a confidence interval?”
To do so, you use a formula that refers to a slider for a field in a statistical object.
1. Create a slider by dragging one from the shelf or choosing Object | New | Slider.
173
Fathom 2 Help
Imagine you’re working with a random sample from some population, such as a sample of voters, a
sample of patients with high blood pressure, or a sample of elementary schools in a state. You measure
something about the sample, for example, the proportion planning to vote for a certain candidate, the
mean blood pressure, or the difference in proportion of students passing a test in two groups. The thing
you measure is an estimate of what is true for the population, in other words, a population parameter.
But the single number is not very useful. Much more useful is the range of values in which you think the
population parameter lies and how confident you are about that range.
Fathom’s interval estimate objects are designed to compute both the estimate and its confidence
interval.
Attribute types Estimate types
174
How To’s
Use this estimate if you have a sample of numeric data for which you want an estimate of and confidence
interval for the mean.
1. Create a new estimate by dragging one from the shelf or choosing Object | New | Interval Estimate.
2. From the pop-up menu in the estimate’s upper-right corner, choose Estimate Mean.
3. Drop the attribute of interest into the prompt in the top section of the estimate, where it says
Attribute (numeric): <unassigned>.
Fathom uses the attribute’s name and values to fill in the estimate at the default confidence level of 95%.
175
Fathom 2 Help
Change the confidence level by double-clicking the 95, typing a new value, and pressing
Return.
See also
Use Sliders to Vary Summary Information in a Statistical Object 173
Shorten the Output in Statistical Inference Objects 173
For numeric data, you need only three numbers, count, mean, and standard deviation, to compute a
confidence interval for a mean. If you have these three numbers, you can type them directly into the
estimate mean object.
1. Create a new estimate by dragging one from the shelf or choosing Object | New | Interval Estimate.
2. From the pop-up menu in the estimate’s upper-right corner, choose Estimate Mean.
3. Double-click the bold, blue field AttributeName in the body of the estimate, and type a reasonable
name.
4. Press Tab to move to the next field and continue inputting appropriate values.
As you enter values, the text changes throughout the estimate object.
This estimate shows a computation of the confidence interval for a sample of 9 scores with a mean of
120 and a standard deviation of 15. Notice that because the estimate is in verbose mode, it describes in
detail the meaning of a confidence interval.
See also
Use Sliders to Vary Summary Information in a Statistical Object 173
Shorten the Output in Statistical Inference Objects 173
176
How To’s
There are two ways to use this estimate, depending on how the data are structured.
The preferred way of structuring data is shown below. Each case is a plant, and the attribute Group tells
whether or not the plant got fertilizer. To assign attributes to the estimate object, the Group attribute,
being categorical, goes to the second line of the attribute pane, and the Height attribute goes to the first
line. This is all shown below. (Notice that the estimate is in terse mode.)
The less preferred way of structuring data is to use one attribute for the values of one group and a
second attribute for the values of the other group. Notice that this means that a single case in the
collection doesn’t really make any sense. The case represents a pair of plants, but there isn’t any good
reason for a particular pair to be assigned to the same case. You can drag either attribute to either line in
the estimate object; the difference will be the first attribute minus the second.
By default, Fathom calculates the difference in means using unpooled variances. Click on the phrase
177
Fathom 2 Help
“unpooled variances” for a pop-up menu that allows you to switch between unpooled variances and
pooled variance. Use pooled variance when you have reason to believe that the standard deviation of the
values is the same for both groups. Unpooled variances use weaker assumptions and produce somewhat
larger intervals than does pooled variance.
You can change the confidence level by double-clicking it and typing a different value or by assigning it a
formula.
You don’t need the raw data to estimate the difference of means. As you can see from the example at
right, you only need to fill in the count, mean, and standard deviation for each of the groups.
Notice here that we’ve used units of Hg*mm for the means and standard deviations. The measurements
are presumably blood pressure in millimeters of mercury.
See also
Enter Summary Statistics in Statistical Inference Objects 172
Use Sliders to Vary Summary Information in a Statistical Object 173
If you have a collection of cases that have a categorical attribute and you want to estimate a proportion
and confidence interval for a categorical value of that attribute, estimate proportion is for you. Say, for
example, you’ve sampled voters in your neighborhood by asking them three questions about an upcoming
election. For each question, you want an estimate of the population proportion that answered “yes.”
1. Create an estimate by dragging one from the shelf or by choosing Object | New | Interval Estimate.
178
How To’s
2. From the pop-up menu in the estimate’s upper-right corner, choose Estimate Proportion.
3. Drop the attribute of interest into the prompt in the top section of the estimate, where it says
Attribute (categorical): <unassigned>.
You can change the category whose proportion is being estimated by clicking on the category name and
choosing from the resulting pop-up menu.
It is common to have only the “tally” of responses to a question and to want to use this sample result to
estimate the proportion for the population. You can do that with the estimate proportion object.
1. Create an estimate by dragging one from the shelf or by choosing Object | New | Interval Estimate.
2. From the pop-up menu in the estimate’s upper-right corner, choose Estimate Proportion.
3. Type in the numbers and names of the category and attribute.
See also
Use Sliders to Vary Summary Information in a Statistical Object 173
Shorten the Output in Statistical Inference Objects 173
Exactly how you do this depends on which of two basic ways your data are structured. First, consider the
preferred data structure. Let’s take a concrete example in which you have a list of 42 employees that
includes their sex and job classification (salaried or hourly). You wish to estimate any difference in
179
Fathom 2 Help
It makes a difference which attribute you assign to each slot. The attribute whose proportions you are
comparing (Sex in this example) goes in the first slot. The attribute that defines the groups (Classification in
this example) goes in the second slot.
3. Drag the two collection attributes to their slots in the estimate object.
The estimate object, in its verbose and terse forms, is shown here.
By default, the estimate was for females rather than males. You can change that by clicking on the word
“female” and choosing male from the pop-up menu.
180
How To’s
Notice that the range of the estimate includes zero, indicating that you would not be justified in
concluding that something other than chance is involved in creating the difference in proportions.
The default confidence level of 95% can be changed by typing or by assigning a formula to use a slider.
See Use Sliders to Vary Summary Information in a Statistical Object 173
Now we’ll consider the less preferred data structure, in which one set of attribute values records the sex
of the salaried employees and another attribute records the sex of the hourly employees. This structure is
less preferred because a case in this collection doesn’t have any meaning; it’s just an arbitrary pairing of
one salaried and one hourly employee.
A portion of the data is shown at right. The blank cells for cases 23 through 26 in the salaried column
make clear that there are different numbers of employees in each group.
In this situation, it doesn’t make any difference which attribute you drop in which slot of the estimate
object.
Once you have dropped the two attributes, the text of the estimate object will fill in with something
nonsensical, similar to what is shown at right. The problem, of course, is that Fathom is assuming that
the attributes conform to the preferred structure.
181
Fathom 2 Help
To fix things, click your mouse on the first part of the phrase and choose the single-word option, in this
case Salaried.
The result of this change, shown at right, is that Fathom treats each attribute as a group and compares the
proportion in one with the proportion in the other.
You can change the category being used for comparison by clicking on the chosen category, in this case
“female”.
You can change the confidence level by typing in a new value or by giving it a formula.
Sometimes you don’t have the raw data from which you wish to estimate a difference of proportions.
Let’s take a concrete example in which you know the counts for sex and job classification (salaried or
hourly) of 42 employees and you wish to estimate any difference in proportion of males between the two
job classifications. The summary data are as follows:
Salaried, female: 10
Salaried, male: 12
Hourly, female: 17
Hourly, male: 13
1. Create a new estimate by dragging one from the shelf or by choosing Object | New | Interval
Estimate.
2. From the pop-up menu in the estimate’s upper-right corner, choose Difference of Proportions.
182
How To’s
3. Double-click the bold, blue field FirstAttribute in the body of the estimate, and type a reasonable
name, such as Salaried.
4. Press Tab to move to the next field, and continue inputting appropriate values.
As you enter values, the text changes throughout the object. The result for this example is shown here.
You can change the confidence level by double-clicking it and typing a different value.
Note that the estimate of the difference of proportions goes from –0.16 to 0.38 at the 95% confidence
level. Because this confidence interval includes zero, we are not justified in saying that the difference of
proportions is greater than zero.
You’re working with a collection that can be treated as a random sample from some population. For
example, it could be a sample of students from a school, data from an experiment about animal learning,
or a sample of census data from a state. You have a hypothesis you wish to test, such as test scores for
students in the school have risen above some threshold, animal learning works better with rewards than it
does with punishments, or median incomes for young workers measured in constant dollars has increased
in the last ten years. These hypotheses are about the population, but your data are for a sample.
The statistical inference process can help you decide whether the measurements in the sample can be
explained by chance variation as opposed to your hypothesis.
Attribute types Test types
One numeric attribute (e.g., height) Test Mean from Raw Data ( 184 t 184 -Test) 184
Two numeric attributes (e.g., income90 and Compare Means from Raw Data (Two-Sample t-
income00) Test) 186
Test Correlation 202
One categorical attribute (e.g., FavoredCandidate Goodness of Fit (Chi-Square) Test from Raw Data
with values “cand1”, “cand2”, and “cand3”) 195
Two categorical attributes (e.g., sex and Test for Independence from Raw Data (Chi-Square
maritalStatus) test) 199
Compare Proportions from Raw Data (Z Test) 191
One numeric and one categorical attribute (e.g., Analysis of Variance (ANOVA) 188
income and sex)
Compare Means from Raw Data (Two-Sample t-
Test) 186 (if the categorical attribute has exactly two
values)
Note: To do a paired t-test, create a new attribute containing the difference between the two elements, and
perform a hypothesis test to see if the mean difference is zero (see Define an Attribute by Formula Using
183
Fathom 2 Help
1. Create an empty hypothesis test by dragging one from the shelf or by choosing Object | New |
Hypothesis Test.
2. Choose the type of test you want from the test’s pop-up menu.
3. Drop attributes on the prompt to have Fathom compute results. In a test, blue text is editable; use it
to control the test (e.g., to change the hypothesized mean; we changed it to 38). Some blue text brings
up a pop-up menu, as shown here (we’re making this a one-tailed test).
If you do not drag attributes to the test, you can still do most tests based on summary statistics. Select the
test from the pop-up menu, then enter the attribute names, sample size, and so on, in the test. (Some
tests, such as ANOVA, require original data, however.)
Once you have an analysis, the Test menu is active. You can use this menu to change to nonverbose, or
terse, mode or to plot a distribution relevant to your test.
184
How To’s
If you have a collection of numeric data (e.g., the lengths of widgets coming off an assembly line) and a
hypothesis about the mean of these values (e.g., their average length has gotten above some threshold),
you can use a hypothesis test object to determine how likely it is that your observed results can be
explained by chance alone even if your hypothesis is not true (e.g., the mean length of widgets is at the
threshold).
1. Drag a hypothesis test from the shelf or choose Object | New | Hypothesis Test.
You don’t need to have the raw data to perform a t-test. All you need is the count, sample mean, and
standard deviation. This example shows a sample of colleges and acceptance rates, testing whether the
population mean acceptance rate is significantly different from 0.5.
1. Drag a hypothesis test from the shelf or choose Object | New | Hypothesis Test then choose Test
Mean in the test’s pop-up menu.
2. Edit the blue text to fill in your information (see Enter Summary Statistics in Statistical Inference
Objects 172 ).
185
Fathom 2 Help
3. Choose the desired sidedness of the test by clicking on “is not equal to” and choosing the option you
want from the pop-up menu.
Any of the editable values can be given a formula instead of a typed value. Particularly useful is using a
formula to tie a value to a slider. See Use Sliders to Vary Summary Information in a Statistical Object 173 .
For example, by using a slider for the sample count, you can discover what sample size would be required
to reach some threshold p-value.
There are two ways to use this test, depending on how the data are structured.
The preferred way of structuring the data is shown here. Each case is a plant, and the attribute Group
tells whether or not the plant got fertilizer. To assign attributes to the estimate object, the Group attribute,
being categorical, goes to the second line of the attribute pane, and the Height attribute goes to the first
line. (Notice that the test is in terse mode.)
The less preferred way of structuring the data is to use one attribute for the values of one group and a
second attribute for the values of the other group. Notice that this means that a single case in the
collection doesn’t really make any sense; it represents a pair of plants, but there isn’t any good reason for
a particular pair to be assigned to the same case. Drop either attribute on either line in the test object; the
difference will be the first attribute minus the second.
186
How To’s
By default, Fathom calculates the significance of the difference in means using unpooled variances. Click
on the phrase “unpooled variances” for a pop-up menu that allows you to switch between unpooled
variances and pooled variance. Use pooled variance when you have reason to believe that the standard
deviation of the values is the same for both groups. Unpooled variances use weaker assumptions and
produce somewhat larger p-values than pooled.
If you have only summary statistics, Fathom can still perform a comparison of means.
1. Make a hypothesis test object by dragging one from the shelf or by choosing Object | New |
Hypothesis Test. Choose Compare Means from the test’s pop-up menu.
2. Edit the blue text with the values for count, mean, and standard deviation for each group.
Notice that you can use values with units. (The values here are millimeters of mercury used to measure
blood pressure.)
Linking values in the test to sliders can be very informative. See Use Sliders to Vary Summary
187
Fathom 2 Help
4.10.6.6 Compare Means When the Values Are Paired (A Paired t-Test)
When you have two sets of values that are paired, using a compare means test would throw away the
information that the values are paired. Instead, construct a new attribute that computes the differences
for each case. Then use a test mean to decide if the difference is significantly different from zero (or
whatever value you hypothesize).
This is shown below for an example in which we have a random sample of student test scores before and
after a unit of instruction.
1. Create a new test by dragging one from the shelf or by choosing Object | New | Hypothesis Test.
2. From the test’s pop-up menu, choose Analysis of Variance.
3. Drop attributes on the prompts at the top of the test.
An example is shown below. It appears that the mean incomes of the 13 races in a sample of 500 people
in Alaska are similar enough that chance alone can explain their differences.
188
How To’s
To plot the F-distribution showing the p-value as a shaded proportion of the area under the curve,
choose Test | Show Test Statistic Distribution.
Suppose you read poll results and you want to know whether they indicate that the proposition you are
hoping will pass is really ahead. In this case, you have the summary data to enter into a test proportion
object.
1. Create a new test by dragging one from the shelf or by choosing Object | New | Hypothesis Test.
2. From the test’s pop-up menu, choose Compare Means.
3. Edit the values in blue to reflect your data.
The example shows that 55% of a random sample of 780 voters in favor of Proposition 1 is a lead that
cannot be explained by chance.
189
Fathom 2 Help
When you have collected data that contain categorical information, one question you can ask is whether
the proportion of cases corresponding to a certain category differs from some particular value. Fathom’s
test proportion object will help you answer this question by computing the probability that the difference
you observe could reasonably happen by chance.
1. Create a new test by dragging one from the shelf or by choosing Object | New | Hypothesis Test.
2. From the test’s pop-up menu, choose Test Proportion.
3. Drop a categorical attribute on the prompt at the top of the object.
This example deals with the situation in which only 28% of the staff in a computer department are
women, while the industry average is 40%. The question is whether something other than chance is
required to explain this difference. The test shows that 11% of the time you would expect to get a
difference this great or greater just by chance
Note that, as usual, you can edit various parts of this test.
A pop-up menu under the category name allows you to change the category whose proportion
is being tested.
A pop-up menu under the phrase describing the direction of the test can be used to change the
direction.
190
How To’s
You can edit by typing or by providing a formula for the test proportion.
Choosing Test | Show p_hat Distribution displays a plot similar to the one at right, with the proportion
of the area under the distribution equal to the test’s p-value.
For this test, Fathom uses the normal approximation to the binomial distribution when n*p ≥ 10 and n*(1
- p) ≥ 10. Under these conditions, it will compute a z-statistic. Otherwise, it will use a binomial statistic.
(The test itself tells you which is used.)
If you have categorical data for individual cases and you want to test whether or not an observed
difference of proportions for two groups is significant, you can use Fathom’s compare proportions test.
Exactly how you do this depends on which of two basic ways your data are structured. First, consider the
preferred data structure. Let’s take a concrete example in which you have a list of 42 employees that
includes their sex and job classification (salaried or hourly). You wish to test whether the observed greater
proportion of males for salaried as compared with hourly is statistically significant.
The portion of data shown at right shows that each case represents an employee with Classification and Sex
as attributes.
1. Create a new hypothesis test by dragging one from the shelf or by choosing Object | New |
Hypothesis Test.
2. From the test’s pop-up menu, choose Compare Proportions.
Look carefully at the top pane of the estimate. Notice, as shown here, that one attribute slot is labeled
“categorical” and the other is labeled “categorical or grouping.”
It makes a difference which attribute you assign to each slot. The attribute whose proportions you are
comparing (Sex in this example) goes in the first slot. The attribute that defines the groups (Classification in
this example) goes in the second slot.
3. Drop the two collection attributes on their slots in the test object.
The resulting test object is shown here.
191
Fathom 2 Help
There are two problems with this test object: We want to compare the proportion of males rather than
the proportion of females; and our alternative hypothesis is that the proportion of males for salaried
works is greater than that of hourly workers.
4. Click on the word “female” and choose male from the pop-up menu.
5. Repeat this change for the other occurrence of “female”.
6. Click on “Sex where Classification is Hourly” and choose Sex where Classification is Salaried
from the pop-up menu.
7. Finally, click on “is not equal to” and change it to is greater than.
Notice that “the probability of getting a value of z this great or greater would be 0.21.” This means that
we cannot rule out chance as an explanation for the observed difference of proportions.
Now we’ll consider the less preferred data structure, in which one attribute records the sex of the salaried
192
How To’s
employees and another attribute records the sex of the hourly employees. This structure is less preferred
because a case in this collection doesn’t have any meaning; it’s just an arbitrary pairing of one salaried and
one hourly employee.
A portion of the data is shown at right. The blank cells for cases 23 through 26 in the salaried column
make clear that there are a different number of employees in each group.
In this situation, it doesn’t make any difference which attribute you drop in which slot of the test object.
Once you have dropped the two attributes, the text of the test object will fill in with something
nonsensical, similar to what is shown. The problem, of course, is that Fathom is assuming that the
attributes conform to the preferred structure.
To fix things, click your mouse on the first part of the phrase and choose the single-word option, in this
case Salaried.
This change results in Fathom treating each attribute as a group and comparing the proportion in one
with the proportion in the other.
Change the category being used for comparison by clicking on the chosen category, in this
case female.
193
Fathom 2 Help
employees and you wish to carry out a hypothesis test of whether or not the greater proportion of males
in the salaried positions, as compared with the hourly positions, is explainable by chance. The summary
data are as follows:
Salaried, female: 10
Salaried, male: 12
Hourly, female: 17
Hourly, male: 13
1. Create a new test by dragging one from the shelf or by choosing Object | New | Hypothesis Test.
2. From the test’s pop-up menu, choose Compare Proportions.
You get a test comparing proportions, with default values; the blue fields are editable.
3. Double-click the bold, blue field FirstAttribute in the body of the estimate, and type a reasonable
name, such as Salaried.
4. Press Tab to move to the next field and continue entering appropriate values.
As you enter values, the text changes throughout the object. The result for our example is shown here.
By default, the test object is configured for a two-tailed test, but in our example, we’re interested
specifically in whether the proportion of males for the salaried jobs is greater than it is for the hourly jobs.
5. Change “is not equal to” by clicking it and choosing is greater than from the pop-up menu.
194
How To’s
Looking at the result, we see that the p-value for this test is 0.21, meaning that about 20% of the time we
would get this level of difference between the two proportions just by chance. Thus, we cannot rule out
chance as the cause of the difference.
It is helpful to get a visual feeling for the p-value.
6. With the test object selected, choose Test | Show Test Statistic Distribution.
The result is the function plot shown here, with the area to the right of the computed value for z (0.7995)
shaded to show the proportion of the time this value or greater would be obtained if the null hypothesis
were true.
When you have data with categorical information, you can use a goodness of fit test to determine whether
or not the distribution of values among the categories conforms to what was expected. The most
common situation is that all categories are expected to be equal.
1. Create a new test by dragging one from the shelf or by choosing Object | New | Hypothesis Test.
2. From the test’s pop-up menu, choose Goodness of Fit.
3. Drop a categorical attribute on the prompt at the top of the object.
195
Fathom 2 Help
As an example, suppose you have rolled a particular die 100 times, with the results shown in this table.
Just by inspection, you suspect that the die is loaded in favor of 6. But couldn’t this happen by chance?
The result shows that, yes, it could happen by chance, but only about two times out of a thousand;
therefore, you are justified in claiming that the die is loaded.
But not every situation tests for equal probability for each category. For example, you are told what the
proportion of different colors of M&M candies is supposed to be. You want to see, given a random sample
of 120, whether those proportions are reasonable.
After making the test and dropping the attribute in the top pane of the test, click on “are not equally
likely” to bring up a pop-up menu. From that menu, choose have probabilities given above. This
creates a second column in the table that, initially, has all equal probabilities. Edit these values to
correspond to the probabilities you are testing against.
In the example at right, the p-value of 0.2 tells us that it would not be unusual to get a set of counts this
different (or more different) when sampling from a population with the given proportions for each color.
Choose Test | Show Test Statistic Distribution to bring up a graph of the chi-square distribution in
which the shaded area corresponds to the probability of getting a chi-square as great or greater than that
observed if the null hypothesis were true.
This plot corresponds to the M&M situation discussed above. The shaded area is 20% of the total area
under the curve.
4.10.6.13 Goodness of Fit (Chi-Square Test) from Summary Data: Equal Probabilities
If you only have the counts for categorical outcomes, you can still use Fathom to do a goodness of fit
test, because you can enter the attribute name, number of categories, category names, and counts into the
test.
196
How To’s
Let’s suppose you are playing a game with a friend using a spinner that supposedly has three equal
outcomes: A, B, and C. But you think the spinner isn’t fair, so you collect the results of a bunch of spins
as tallies for A, B, and C. You get 24 for A, 37 for B, and 20 for C.
1. Create a new test by dragging one from the shelf or by choosing Object | New | Hypothesis Test.
2. From the test’s pop-up menu, choose Goodness of Fit.
3. Double-click one of the blue <AttributeName> prompts in the body of the test and type Spin.
4. Type 3 in place of 0 for number of categories. (Notice that you now have a table in which you can fill
in information.)
5. Click on each of the RowCategory cells successively and change them to “A,” “B,” and “C.”
6. Fill in the three counts next to their proper category.
You should have something similar to the test shown at right.
With a p-value of about 0.05, you can’t be sure that the results are not simply due to chance. Perhaps a
bigger sample will lead to a significant result!
197
Fathom 2 Help
4.10.6.14 Goodness of Fit (Chi-Square Test) from Summary Data: Unequal Probabilities
There are certainly many situations in which categorical outcomes are not expected to be equally likely.
Here we’ll consider flipping a pair of coins 100 times and keeping track of the number of times you get 0
heads, 1 head, and 2 heads. Theoretically, these outcomes have probabilities 0.25, 0.50, and 0.25,
respectively.
1. Create a new test by dragging one from the shelf or by choosing Object | New | Hypothesis Test.
2. From the pop-up menu in the test’s upper-right corner, choose Goodness of Fit.
3. Edit the attribute name (in blue) to be Heads.
4. Enter the number of categories (in this case 3).
The table will create three editable rows.
5. Change the row category names from their defaults to “0,” “1,” and “2.”
Notice that the test is, so far, assuming equal likelihood.
6. Click on the phrase “are equally likely” and choose have probabilities given above from the pop-
up menu.
Notice that the table gets a column for probabilities.
7. Enter the appropriate probabilities for each row.
The chi-square statistic and associated p-value are displayed below the table.
With the test object selected, choose Test | Show Test Statistic Distribution to display a plot of the
chi-square distribution with shading to show the probability of getting a statistic as large or larger than the
one observed.
198
How To’s
Suppose you have two categorical attributes and you want to know if there is any relationship between
the two. For example, a school has 70 teachers in three divisions: lower, middle, and upper. You want to
know whether or not a teacher’s assignment by division is independent of sex.
1. Create a new test by dragging one from the shelf or by choosing Object | New | Hypothesis Test.
2. From the test’s pop-up menu, choose Test for Independence.
3. Drop each of the attributes on the prompt of the test.
The results for the Teachers example are shown here (in nonverbose mode). It looks like there may be
some discrimination going on!
The test object computes what the count would be for each cell if the two attributes were independent. These
expected counts are shown in parentheses. When the expected counts in cells drop below five, the
goodness of fit test is no longer very accurate. Fathom warns you about this.
199
Fathom 2 Help
Nothing about doing a test for independence requires you to have raw data. All you need are the tallies
for each possible outcome.
1. Create a new test by dragging one from the shelf or by choosing Object | New | Hypothesis Test.
2. From the test’s the pop-up menu, choose Test for Independence.
3. Edit the values in blue to reflect your data.
Let’s suppose you do a biology experiment with pill bugs in which you divide up the bugs’ living area into
six regions according to temperature and humidity. On one occasion, you count the number of pill bugs
in each region and enter the counts into a Test for Independence, as shown here. The entry of the
number of categories is what expands the table and allows you to rename categories and enter counts in
cells.
By choosing Test | Show Test Statistic Distribution, you can show a graph of the chi-square
distribution. This graph has shading to indicate the probability of getting a value as great or greater than
the observed value if the attributes are actually independent.
200
How To’s
There are many approaches to investigations of a possible relationship between two numeric attributes.
One question that can arise is whether the slope of the observed least-squares regression line is
significantly different from zero. Shown below is an example “playground” that has a collection with only
4 cases. A scatter plot shows the value of attributes x and y, and a test slope object shows (in verbose
mode) the determination of how likely the observed slope would be if there were no relationship between
the two attributes.
1. Create a hypothesis test object by dragging one from the shelf or by choosing Object | New |
Hypothesis Test.
2. Choose Test Slope from the test’s pop-up menu.
3. Drop numeric attributes on the prompt at the top of the test.
A plot of the t-distribution shows the p-value as a shaded area. (Choose Test | Show Test Statistic
Distribution.)
Dragging the points around on the scatter plot to get a feel for what makes a slope significant can be
quite instructive.
Note that this test requires raw data; that is, it cannot be used with summary data.
201
Fathom 2 Help
Conceptually, there is a difference between testing whether a least-squares slope is different from zero
and testing whether a correlation coefficient is different from zero. The first assumes that a response
attribute’s values are being predicted from a predictor’s values. But in a test for correlation, the two
attributes play a completely symmetric role.
1. Create a hypothesis test object by dragging one from the shelf or by choosing Object | New |
Hypothesis Test.
2. Choose Test Correlation from the test’s pop-up menu.
3. Drop numeric attributes on the prompt at the top of the test.
Because a test of correlation only requires two numbers, the count and the correlation coefficient, you
can enter these numbers directly in the test without needing raw data.
In the example, the same “play” data from the description of Test Slope is used to illustrate Test
Correlation. If you construct these two tests side-by-side using the same data, you’ll see that student’s t is
the same for both. This means that if the slope is significant, so is the correlation coefficient.
We build models to understand things in the real world and to make predictions about them (thus, the
crystal ball). The linear models that Fathom provides as statistical objects are geared to work with
continuous attributes under the assumption that there is a linear relationship between a response attribute
and one or more predictor attributes.
The simple regression model is actually a subset of the multiple regression model; that is, you don’t
need to put more than one predictor attribute in a multiple regression model, and if the model only has
one attribute, it’s equivalent to a simple regression model.
202
How To’s
Suppose you have data for two attributes and you wish to predict one from the other. For example, you
might have a sample of airplane types and data about the cost per hour to run them and their range in
miles. Can you predict cost per hour from the range?
1. Create an empty linear model by dragging one from the shelf or by choosing Object | New | Linear
Model.
2. From the model’s pop-up menu, choose Simple Regression.
3. Drop the response attribute on the first line of the top pane.
4. Drop the predictor attribute on the second line of the top pane.
The result for the airplanes example is shown at right. The blue values are editable either by typing in a
new value or by providing a formula. A formula tied to a slider is especially useful. See Use Sliders to Vary
Summary Information in a Statistical Object 173 .
The last paragraph of the model allows you to type in your own value for the predictor to get an estimate
for the response. The confidence level for this estimate is the same as is given for the slope.
You can make a plot of residuals by making a scatter plot of the two attributes. See Make a Residual Plot
58 .
You can also make attributes for predicted values and residuals in the collection using the linRegrPredicted
and linRegrResiduals functions. (see the Reference section on Statistical Functions for Two Attributes 420
.)
203
Fathom 2 Help
To make this explanation concrete, we’ll use the Airplanes.ftm sample document that comes with
Fathom. This document contains a collection of 21 airplanes, with the six attributes shown in the
inspector at right. We would like to be able to predict the cost per hour of operating an airplane based on
some linear combination of the four other numeric attributes.
For now, we’re going to concentrate on the mechanics of building the model, without explaining much
about the underlying statistics or attempting to show good practice. You’ll have to get this from a book
or course.
1. Drag a model object off the shelf or choose Object | New | Linear Model.
204
How To’s
You will almost certainly need to resize the model so you can see all of its parts. It should look similar to
the following illustration:
Notice the regression equation in the second pane from the top. Even without any predictors, we are not
completely in the dark; we can use the mean of the predicted values, and that’s what the equation says.
Also notice the empty horizontal bar. The fact that it is empty represents the fact that 100% of the
variation from the mean is, as yet, unexplained. Our goal is to fill in that bar.
4. Drag the seats attribute into the “Drop attributes here to add predictors to the model” line above the
regression equation.
5. Drag the speed attribute so that it fits just under seats.
You should have something like this:
205
Fathom 2 Help
Each row in the table in the top indicates the incremental effect of adding this predictor to the model last,
that is, after all of the other predictors. Conversely, it can be viewed as the contribution that would be lost
if this predictor were removed from the model. These results will change as predictors are added and
removed from the model, but not when the order of predictors is changed.
Each section of the ribbon in the ribbon chart corresponds to the incremental effect of adding its
predictor in its current position. Changing the order of the predictors will change the graph, but adding or
removing predictors to the model won’t affect the contributions of predictors that precede the change.
By default the ANOVA table at the bottom shows the portion of the variance accounted for by the
model. To show the incremental effect of adding each predictor to the model, choose Model | Show
Predictors In ANOVA Table.
The final section of the multiple regression model contains results for the model as a whole.
There’s a lot going on here! There are two things to focus on before we go into the details: There is now
a regression equation with three terms that can be used to predict costph from seats and speed. Also, the bar
shows that 80% of the variation is accounted for by the linear relationship.
The constant term and the coefficients for the two linear terms. These are also
Coefficient
in the regression equation.
206
How To’s
These numbers are the standard deviation of the sampling distribution of each
Std Error
of the corresponding coefficients.
The probability that a value for the coefficient this different from zero would
p-Value
be obtained if the population value of the coefficient were zero.
The change in R2 that this attribute would contribute if it were the last attribute
DR2 added (i.e., at the bottom of the list).
Regression The equation with which you can predict the response. The “hat” over the
Equation response attribute name indicates that it is a predicted value.
The bar represents the variation in the response. Each segment of the ribbon
Ribbon Chart represents the portion of that variation “accounted” for by an attribute. The
blank segment represents the remaining, unaccounted for, variation.
One important number in this table is the p-Value. If this value is small (and a
value of zero means that it’s really small), then it is unlikely that the amount of
ANOVA Table
explanation provided by the model would occur by chance under conditions in
which there was no linear relationship among the attributes.
This is the total R2 value for the model. You can think of it as the proportion
R-Squared
of variation in the response accounted for by the model.
4.10.7.2.3 Adding, Removing, and Moving Predictor Attributes: The Goal of Model Building
You can drag additional attributes to the table of attributes. Notice that as you drag, you can decide where
in the table an additional attribute goes, as signified by the location of the thick black line. (You can drag
more than one attribute at a time to the table.)
You can remove an attribute by selecting its line in the table of attributes and then choosing Test |
Remove Attribute.
You can move an attribute within the table by dragging it to a new location. Notice that the portion of
the ribbon chart corresponding to an attribute turns red when you select the attribute. Moving the
attribute changes its place in the table, the regression equation, and the ribbon chart. Notice that moving
things around does not change the coefficient for any of the attributes or the total R2. It does, however,
change the sequential contribution.
You might well ask, “What is the goal of all this adding, removing, and moving of attributes? How do I
know when I’m done?” Building a multiple regression model is a bit of an art. There’s no single right
answer. To some extent, you’re trying to account for as much of the variation as possible. But you also
want the model to be as simple as possible, and you want the attributes to make sense as you interpret the
model.
Think again about the Airplanes model. Doesn’t it seem likely that fuelgph would be better at explaining
cost than any of the other three possible predictor attributes? In fact, the model with fuelgph as the only
predictor accounts for more of the variation than the model with all of the other three! Simple linear
regression turns out to be simpler and better.
207
Fathom 2 Help
You may not need all of the multiple regression object to show. Two commands in the Test menu hide
portions of the object. Hide Sequential Contributions Chart hides the ribbon chart, and Hide ANOVA
Table hides the bottom table. Also, unchecking Verbose will get rid of a few words. Using all these
options, the multiple regression model of Airplanes looks like the example here.
You can choose Test | Show Predictors in ANOVA Table to examine the intermediate statistics
computed for each predictor, as shown below.
Of course, you want to take a look at the residuals. Once you have a multiple regression model, this is
easy.
1. With the model selected, choose Edit | Inspect Linear Model.
2. Drag attributes from the Cases panel of this inspector to a graph.
208
How To’s
Here’s a plot of residuals versus predicted values for the Airplanes data.
You can use this inspector to define new attributes in terms of the predicted values and residuals. You
can also copy the formulas, especially the formula for the predicted value, and use them elsewhere. (Of
course, the formulas won’t dynamically update.)
You can make a case table that includes the predicted values and residuals by first selecting the model
object and making a new case table, or connect the model to an existing case table by dragging the model
name from the upper-left corner of the model object to the case table.
Suppose you have some bivariate data and you have some reason to believe that the right model for the
data is a third-degree polynomial. You can use the multiple regression model to find the coefficients for
each term.
At right are some fictitious data that look vaguely cubic. The trick is to realize that by adding attributes
for xSquared and xCubed, you can reduce the problem of polynomial regression to multiple linear
regression.
1. Drag a model object from the shelf or choose Object | New | Linear Model and choose Multiple
Regression from its pop-up menu.
2. Make attributes for powers of x.
209
Fathom 2 Help
Now build up the linear model as shown here. Notice in the ribbon chart that each term contributes
significantly to the model.
Finally, show the model’s inspector, copy the regression equation, and plot it on top of the original scatter
plot.
Note: You have to edit the equation, replacing xSquared and xCubed with x^2 and x^3 before the function
will plot.
The resulting plot, with the cubic on top of the data, is shown here.
210
How To’s
As usual with a measures collection, you can collect more measures and control the number and
conditions for collecting measures in the Collect Measures panel of the collection’s inspector. If the
attributes in the test or estimate are randomly generated, Fathom rerandomizes before collecting each set
of measures.
211
Fathom 2 Help
Use space wisely. Delete unnecessary objects, objects that don’t help you tell your story. Move
related objects together. (For example, put the collection and graph of the collection that
serves as a population in one area, the sample collection and its objects in another, and the
measures collection and its objects in another.)
Subsections
Add Comments to a Collection 212
Work with Text in Text Objects 213
Paste Pictures 213
See also
Change the Appearance of Cases in a Collection 106
Each Fathom collection has a special place for holding documentation: the Comments panel. The
biggest advantage of the Comments panel over text objects is that many users read text objects and then
delete them to free up space; the Comments stay with the collection.
Note: When you import data, text that Fathom recognizes as not being the data itself is stored in the
collection’s comments automatically.
1. Double-click the collection (the box of balls) to show its inspector.
2. Click the Comments tab to go to the Comments panel.
3. Type or paste comments into the pane. Fathom will store them when you save the file.
See also
Work with Text in Text Objects 213
Format Text in Text Objects 213
Paste Pictures 213
212
How To’s
3. If a symbol you want isn’t visible, the rightmost button of the symbols palette is a pop-up menu with
even more symbols.
213
Fathom 2 Help
2. Switch to Fathom, making sure that nothing is selected in your document. (Click in an empty space to
deselect all objects.)
3. Choose Edit | Paste Picture.
You can resize the picture just as you would any other object in Fathom, by dragging on the edges of its
frame.
You can also copy and paste pictures of Fathom objects in a Fathom document (for example, if you want
several static graphs of randomizations).
1. Select the object and choose Edit | Copy As Picture.
2. Click in an empty area of the Fathom document (to deselect all objects).
3. Choose Edit | Paste Picture.
214
How To’s
5 Movies
This is a set of QuickTime movies, each around a minute or two long. They show how to use some of
Fathom’s features. To see and hear them, you need the free QuickTime player. (It comes already installed
on Mac. Windows users can download installers from Apple.) If you prefer, you can watch these movies
by browsing the FathomHelp/Movies folder in the Fathom application folder on your hard disk.
Subsections
Graphs 215
Importing Data 217
Inspectors 218
Measures 219
Menus 219
Plotting 220
Summary Tables 223
Surveys 224
Using Formulas 225
Other 226
5.1 Graphs
Click the movie's name to open the movie.
215
Fathom 2 Help
Legend Attributes
Learn how to split plots by
categorical and numeric attributes
to create a legend that gives each
point its own symbol or color. This
splitting is often helpful in looking
for differences between groups or
in spotting trends. Legend Attributes
216
Movies
Rescaling Axes
Rescaling Axes
Zooming in Graphs
Zooming in Graphs
217
Fathom 2 Help
5.3 Inspectors
218
Movies
5.4 Measures
5.5 Menus
219
Fathom 2 Help
5.6 Plotting
220
Movies
221
Fathom 2 Help
222
Movies
223
Fathom 2 Help
5.8 Surveys
Creating and Publishing a Survey
Getting Results
224
Movies
Getting Results
Uploading by Students
Uploading by Students
225
Fathom 2 Help
5.10 Other
Selecting Cases
Selecting Cases
Taking a Sample
226
Movies
Taking a Sample
6 Tours
The tours introduce you to all the features of Fathom in different contexts. They’re written for you to do
in order and step-by-step. We expect you to use a computer running Fathom while you go through the
tours. You may, of course, use the tours differently—jumping from topic to topic, for example—but
keep in mind that you may come across instructions in the wrong order. That is, in later tours, you will
need to know basics covered in the first tours.
Unless you’re very comfortable with computers and with statistical concepts, you probably don’t want to
do the tours in one run. Consider doing a couple of tours, then using Fathom for a while to get used to
the basics before doing later tours.
From time to time, we ask you to open a file. These files are in the Sample Documents | Learning
Guide Starters folder in the Fathom directory.
Although Fathom has many keyboard shortcuts, they are not mentioned in these tours, nor are Fathom’s
context (right-click) menus. If you prefer keyboard shortcuts to choosing from menus, you might want to
keep the Quick Reference handy as you use these tours. If you like context menus, you can usually get
what you want from them. All Fathom objects have their own context menus that contain most of the
commands that apply to each.
Subsections
227
Fathom 2 Help
228
Tours
The document currently has two objects: a text object, explaining where the data come from, and a
collection, which is where the data are stored. Each gold ball in the collection represents one case, which
in this file is one person.
5. Put your cursor over the collection.
The status bar at the bottom left of the Fathom window gives some information about the data: This
collection has 500 cases and 19 attributes, or variables.
229
Fathom 2 Help
Graphing Data
There are many different relationships to explore, but first let’s see where all of these students are from.
8. Drag a graph object from the object shelf into your document, or
choose Object | New |Graph. You get an empty graph because you
haven’t yet told Fathom what attribute to graph.
9. Drag the attribute name Place from the inspector, and drop it on the
prompt below the graph’s horizontal axis. (You could, instead, have
dropped it on the vertical axis.)
By default, you get a bar chart showing where the students live and how many live in each place. The
United Kingdom has the most students in this sample, and Queensland (a state in Australia) has the
fewest.
To see exactly how many students are from South Africa, hold the
cursor over the bar representing the South African students, and read
the status bar in the lower-left corner of the Fathom window.
The status bar indicates what percentage of students come from each
place. You can also show this information on the graph.
10. Use the pop-up menu at the top right of the graph to change the
graph to a ribbon chart.
The data appear as a single big block, representing all of the
values, divided into sections; the axis is labeled with percentages
230
Tours
Continuing our exploration of the data one attribute at a time, let’s see
how old the students are.
11. Drag Age from the inspector, and drop it below the horizontal axis
to replace Place.
Now each student is represented by a dot placed along the axis.
12. Select the oldest student by clicking their dot in the dot plot.
Where is that student from? Notice that when you select a data point,
the inspector shows that case. You can read all of that student’s values
(this student, female, walks to school
and likes English).
Is there any relationship between where students are from and their age? One way to look for
relationships is to make use of the fact that selection in one graph is reflected in another.
13. Make a second graph next to the first one and drag Place to its horizontal axis.
14. Select the bar representing Queensland students by clicking it.
The bar turns red to show the selection of students. Notice that some of the dots in the dot plot have
also turned red and that the students from Queensland are highlighted in the collection.
This graph shows, for example, that the narrowest range of ages is for the Queensland students; the
South African students span the greatest range and have all the oldest students; and the youngest students
231
Fathom 2 Help
20. Move the inspector to the right of the Fathom window by dragging its title bar.
21. Delete the Place graph.
22. Move the Age graph to the left side of the window by dragging the top of its frame.
23. Finally, simplify the Age graph by removing Place. Select the graph and choose Graph | Remove Y
Attribute.
232
Tours
Exploring Graphs
We saw how to use selection to explore relationships. Fathom also allows us to drag things in graphs. We
can use this to further explore the Age graph.
24. Duplicate the Age graph by selecting it and choosing Object | Duplicate Graph.
25. Change one of the graphs into a histogram.
26. Select one of the histogram bars.
This selects a strip of dots in the dot plot. Put the cursor over the selected bar to read about it in the
status area: the range of values it contains and how many cases are in it.
27. In the histogram, drag an edge of one of the bins to make the
bins wider.
Each bin now represents more cases, so some of the bins are
too tall to see their tops.
233
Fathom 2 Help
The dots are now in bins, each the same width. Sometimes the
choice of bin width affects the overall look of the data, for
example, whether there’s one tallest bin or more than one. Let’s
dynamically explore how bin width affects the histogram’s look.
28. Drag the 70 on the vertical axis down until you can see the
tops of all the bins.
Notice how some of the bumpiness in the histogram has been
washed out with the wider bins. Selecting a bin now selects a
broader strip of dots in the dot plot.
Although there are other ways of rescaling graph axes (see Fathom Help for details), the easiest is simply
to drag the axes: Dragging near the ends of axes dilates them, dragging the middle shifts them.
29. Choose Graph | Rescale Graph Axes to see all the data.
Going Further
Using the techniques you’ve learned, answer these questions or make up your own.
Is math more popular with males or females?
Which place has the highest proportion of students with Internet access?
Which country’s students have to travel farthest to get to school?
Which place has the highest proportion of students who walk to school? Are these the students who
have the greatest travel time?
Entering Data
For this tour, we’ll be entering data from scratch. For this, we need an empty case table.
1. If Fathom isn’t running, launch it. Otherwise, create a new, empty document by choosing File | New.
(You might want to maximize your window to give yourself plenty of room to work.)
234
Tours
3. Click <new>.
The Tab key moves you to the next cell. (If you’re at the end of a case, it moves you to the first cell of the
next case, so you can tab your way through the whole table. See Fathom Help: Keyboard Shortcuts 441
for more on moving within and between cells.)
11. Type 76.5 in for Scott’s height in inches, and press Tab.
Because this value has a unit, a new row (the units row) appears above the first case, and the unit
name appears in its cell in green. This is Fathom’s way of indicating that it recognizes the unit as a unit
and that the unit applies to the entire attribute.
12. Type in the rest of the data for Scott, and press Tab to move to the first cell of the next case.
After typing inches for each of the first cells, you don’t have to retype inches, as you’ll see in the next case.
235
Fathom 2 Help
13. Type the rest of the people’s names and their measurements (just the numbers). As soon as you leave
each cell that has a unit applied, Fathom inserts the unit’s abbreviation.
14. Save the document to preserve your data.
We can add a line to the graph to help quantify the relationship between wrist size and height.
20. Select the graph.
Notice that now the Graph menu appears and the Table menu is gone. If you can’t find a menu or a
command for an object, try selecting the object to make its menu appear.
236
Tours
22. Run the cursor along this line, noticing how the cursor changes
shape. Experiment with using different cursors to drag the
line.
23. Move the line to give a pretty good fit to the data by first dragging one end and then the other.
The line is an eyeball fit; different people will fit the line somewhat differently. Fathom can also compute
several standard lines of fit. One of these is the least-squares line. Let’s find out just what that means.
24. With the graph selected (so the Graph menu is in the menu
bar), choose Graph | Show Squares.
25. Drag the line around. Watch the squares and their sum change.
26. Adjust the line to make the sum of the squares as small as you can make it.
The line that minimizes the sum of the squares of the residual is the least-squares line.
27. Choose Graph | Least-Squares Line.
A second line appears in the graph. This one is computed from the data and is not movable. Were
you close?
28. Choose Graph | Remove Movable Line.
What does the equation of the least-squares line tell us about the relationship between wrist size and arm
span? (Someone with a wrist 1 inch bigger around than another person will likely have an arm span about
14 inches longer.)
Dragging Data
In Fathom, we can drag data within graphs to change their values. This allows us to explore how changes
in the data will change our analysis
237
Fathom 2 Help
Drag data points around in the graph, noticing how much effect each point has on the line’s equation.
After exploring the effects of changing the data, we could undo all the moves until we have the original
data back—but we’ll use a different method.
30. Select the collection and choose File | Revert Collection.
After warning you that this step is not undoable, Fathom puts all the data back in its original state (the
state at which you last saved) without taking away the other changes you made (the graph and its line,
for example).
One way to judge how well a line fits data is to look at a residual plot.
31. Select the graph and choose Graph | Make Residual Plot.
238
Tours
34. Make a new graph with WristCircum on the horizontal axis and
Height on the vertical axis. Add a least-squares line.
35. Measure or estimate your wrist circumference
in inches.
36. Click and hold your mouse over the line. The cursor changes
and a red dot appears on the line.
37. With the red dot showing, drag the cursor around in the
graph. The dot traces the line, and its coordinates appear in
the graph.
38. Use this feature to predict your height. How far off is the
prediction from your actual height?
Although you can’t measure your own arm span without help, maybe we can use this technique to predict
it. Let’s look at the relationships of the other attributes to arm span to see which is best for predicting it.
Look at the graph of ArmSpan versus WristCircum. After the equation of the least-squares line, the r2 value
is given. The closer this number is to 1, the better the correlation between the two attributes.
We now want to graph arm span against each of the other attributes. Here’s a quick way.
39. Select the ArmSpan graph and choose Object | Duplicate Graph. You get a copy of your graph, with
arm span and the least-squares line in place.
40. Replace WristCircum with Forearm by dropping Forearm below the horizontal axis.
Do forearm length and arm span seem correlated?
41. Select the ArmSpan graph and choose Object | Duplicate Graph. Change the graph to show arm
span as a function of height.
Height and arm span are very highly correlated, aren’t they? This graph will probably give a better
prediction than the Forearm graph.
42. Trace the line to see what value this data predicts for your arm span. Look at the residual plot to see
how far off this is likely to be. (With help, you could see how close this prediction is to your arm
span.)
Going Further
In this tour, we changed how cases are captioned. Look in Fathom Help to learn how to change the
captions to give a whole sentence about each person (using the concat function). Change the gold ball
images in the collection to look like people and base their size and placement on the data values. (See
Fathom Help: Change the Appearance of Cases in a Collection 106 .)
Leonardo da Vinci documented the correlation between height and arm span, as well as correlations
between other human body measurements. You could research what correlations he found and collect
data on friends and family to verify them.
239
Fathom 2 Help
Skim through the list of attributes, clicking on headings of interest in the left pane, and reading about the
attributes in the right pane. For now, we’ll keep the default request but add one more attribute.
6. In Education, check School Attendance.
7. Click Download Data.
240
Tours
Fathom connects to the Internet and submits your request to IPUMS (Integrated Public Use Microdata
Series, at the University of Minnesota), which has a searchable database of census microdata samples.
Fathom decodes and imports the results into a collection. (If left coded, all data would be in the form of
numbers, rather than, for example, “male” and “female.”)
To compare racial diversity, we’d like to quantify it. We’ll do this in two ways. First, we’ll look at the
proportion of the majority race (the smaller that proportion is, the more racially diverse an area is), and
then we’ll look at how many different racial groups live in an area. We’ll use a summary table.
10. Drag a summary table from the shelf.
11. Drag Race_General to the summary table and drop it on the
down arrow that appears when you’re over the drop area.
Right now, we’re investigating the majority race, so we’re looking at the proportion of whites. We can
make a new summary table to calculate only this proportion.
241
Fathom 2 Help
14. Make a new summary table, and drop the collection’s name in
it.
You get the count of the number of cases in the collection.
The collection has been connected to the summary table, so it “knows” about the collection and its
attributes. We can edit the formula to calculate only the proportion of cases whose race is white.
15. Double-click the formula to show the formula editor, and delete the existing formula.
We could type the formula we want, but, instead, we’ll use the attribute and function list in the formula
editor itself.
16. Open the Functions list in the middle right pane of the
formula editor. (You can resize the panes by dragging their
edges or resize the editor by dragging its lower-right corner.)
17. Open the list of Statistical functions, then open One Attribute.
18. Scroll down to proportion, and double-click it to insert it in
the formula.
Although you can type everything into the formula editor,
sometimes it’s easier (and more accurate) to choose items from
the list (for example, if you don’t remember exactly what you
named a particular attribute or if its name is long).
Now we want Fathom to calculate how many distinct racial types are in this sample.
20. Choose Summary | Add Formula (the menu won’t appear unless the summary table is selected).
21. Enter the formula uniqueValues(Race_General) and close the
editor.
Your numbers may be a bit different from those shown here. When IPUMS has more cases available than
we are asking for, we get a simple random sample. Try downloading data several times to see how much
the numbers change.
We now have a rough idea of the racial diversity for the United States as a whole. We want to look at how
242
Tours
the diversity varies around the country. When we ask for different data, our data will be replaced, so we
need a record of the values we got for the country as a whole. We can make a picture of this table and
keep it for future reference.
22. Make the summary table a good size—as small as possible but still showing all the information.
23. Select the summary table and choose Edit | Copy As Picture.
24. Click in a blank place in the document to deselect the table.
25. Choose Edit | Paste Picture.
This isn’t a live summary table and won’t change when we change the data. (You might also want to have
a picture of the graph of race.)
School Attendance
You could continue the exploration of racial diversity, looking at different states and metropolitan areas
to find the most and least racially diverse places in the United States. But let’s move on and look at some
of the other attributes.
243
Fathom 2 Help
Who’s the oldest person in school now? What does N/A mean
for this attribute?
Let’s look at the age range of students and the rate of school
attendance over the years. The IPUMS collection has samples of
data going back to 1850. It lacks data from the 1890 census (the
data were lost in a fire, unfortunately) and, as of this writing, the
1930 data (which IPUMS will be adding soon).
244
Tours
40. Make a new graph with Age on the horizontal axis and
Census_year on the vertical axis.
The people in school should be selected in this graph, too.
41. Change the graph to a histogram.
What do these graphs tell you about schooling over time in the
state you chose?
Adult Education
Our histogram showed something about the age range of those in school. We can use filtering to look
more closely at the schooling of children or of adults. First let’s look at just the children.
42. Select the ribbon chart and choose Object | Add Filter.
The formula editor appears.
The formula entered for the filter tells Fathom what cases to keep in
the graph.
43. Type the formula age < 18 and click OK.
The ribbon chart changes to show only the
children, and the filter formula appears below
the graph. Notice that none of the other
objects are filtered; the filter applies to this
graph only.
There don’t seem to be any N/A cases for the
earlier years. What might this tell us about the
census for those years?
If we want all our objects to be filtered, we need
to apply the filter to the collection itself. Let's
look at the pattern for adult education.
245
Fathom 2 Help
Going Further
Try downloading data for California in the years 1850 and 2000. Make a ribbon chart with Census_year
on the horizontal axis and Sex in the middle. What do you see? What’s going on here? (Hint: Who came
to California before 1850? Note: The 1850 census did not count Native Americans.) Verify your
hypothesis by getting more data, such as Occupation. How long did it take for the sexes to even out?
(Find out by getting some years in between.)
Explore the idea that people now move around more than did people in the past. (Make an attribute
with a formula that compares people’s current state with their birthplace, such as
if(Birthplace_General = State_FIPS_code), “Same State”, “Moved”
246
Tours
In Fathom, you can use formulas to turn the collection itself into a graphical representation of your data.
4. Double-click the collection to show the inspector.
5. Go to the Display panel by clicking its tab.
There are many special display attributes with formulas in them:
x and y control each icon’s placement horizontally and vertically
(with the origin being the top left corner); image controls the
picture (by default, the familiar gold ball); width and height control
image size (you can see their entire formulas without resizing—
the formulas are simply the planets’ radii); and caption controls
what text appears below each icon.
247
Fathom 2 Help
13. With the case table selected, choose Table | Show Formulas.
A new row appears; the formula cells are shaded to show that
none of the existing attributes are defined by formula.
14. Add a new attribute named Volume.
15. Double-click its formula cell to show the formula editor.
The top of the formula editor holds the formula itself. You can
type into the formula pane, or use the keypad and attribute and
function lists to enter things into the formula.
19. Either type radius (the formula editor is not case sensitive), or expand the Attributes list and double-
click Radius.
When you enter a valid attribute name, it turns magenta.
20. Type ^ to make an exponent, and 3 to cube the radius.
21. Press Enter or click OK to accept the formula and close the formula editor.
The formula appears in the cell, and the attribute values have a gray background to show that they are
computed.
22. Make another new attribute, Density, and give it this formula:
The numerator is the mass in Earth masses, and the denominator is the volume in Earth volumes, so the
density comes out in Earth densities.
23. Graph Density.
24. Click a data point to see which planet it represents.
248
Tours
25. Select more than one data point by dragging a rectangle around
a group of points.
Playing Kepler
Now we’ll study one of the most famous relationships in physics - how the period of an orbit depends on
the orbit’s radius. This relationship was discovered by Johannes Kepler in 1618.
26. Replace Density with orbit_AU, and put year on the vertical axis of the graph. You should see the
points on a gentle upward curve.
It looks like we could predict how long a planet takes to go around the sun if we knew the radius of its
orbit. Let’s try fitting a function.
27. With the graph selected, choose Graph | Plot Function.
28. In the formula editor, enter a plausible function, such as
orbit_AU2.
A curve appears in the graph, and its equation appears below.
The curve is too steep. It looks like we need an exponent between
1 and 2. We’ll use a slider to vary the exponent until it’s right.
The exponent we need is between 1 and 2, so let’s expand the slider’s axis to get finer control.
249
Fathom 2 Help
31. Drag numbers off the slider’s axis until the slider goes roughly
from 1 to 2.
Now we need to get this slider into our function.
32. Double-click the function’s formula below the plot area of the
graph to show its editor.
33. Replace the 2 with the slider’s name (you can find the name in
the Global Values list, or simply type it), and close the editor.
34. Drag the slider’s thumb, and watch the plotted function change
to reflect the slider’s value.
In the graph, there is a gap between Mars and Jupiter. What goes there?
250
Tours
Orbits in Detail
So far, we have been looking at the orbit_AU attribute, which has each planet’s average distance from the
sun. However, planets move in ellipses, not circular orbits. The collection has attributes for the minimum
and maximum orbital distances for each planet: periAU and apheAU (short for perihelion and aphelion).
We can add these attributes to our graph of year versus orbit_AU.
40. Drag the attribute periAU to the horizontal axis of the graph,
but don’t release it yet.
Notice that when the attribute is over the graph, each axis has a
big, gray plus sign. When you move the cursor over it, the
highlighting disappears from the axis, and the plus itself
becomes highlighted.
41. Drop the attribute on the plus.
The attribute is added to the axis: a second set of data points
appears (blue squares rather than gray dots), and a legend
appears at the bottom of the graph, telling you which kind of
dot represents which attribute.
251
Fathom 2 Help
252
Tours
Cumulative Sum
Let’s make a new attribute, C, that is the cumulative total of A for each case.
17. Add the attribute C, giving it the formula A + prev(C).
Note that it’s not A + prev(A), because that would add only two values of A rather than adding to the
previous cumulative sum of A.
18. Make a line plot of C. What shape is this? Why? (There’s a calculus connection here: B is the
derivative of A; C is the integral.)
Fibonacci Numbers
Now that we’re used to how the prev function works in Fathom (there’s also a next function that takes the
value of the next case), we can look at other uses, such as the Fibonacci numbers (1, 1, 2, 3, 5, 8, …). For
this, we’ll start over, making another collection.
253
Fathom 2 Help
19. Scroll down in the Fathom window to give yourself a blank area.
20. Make a new, empty case table with an attribute called Fib.
We need to use a formula that will result in the first two cases getting the value 1 and the rest of the cases
getting the sum of the previous two numbers. For this, we need an if-statement.
21. Select Fib and choose Edit | Edit Formula.
22. Type if(
When you type the open parenthesis, Fathom gives you a template for filling out the if statement. You
want the first two cases to have the value 1.
23. Inside the parentheses, type caseIndex£2. (To get the £ press Ctrl (Win) Option (Mac) and click the <
button that appears on the keypad.)
24. Press Tab to get to the top result clause. (This specifies what value the cases that meet the condition
get.)
25. Type 1 and tab to the next clause.
For all cases that don’t meet the specified condition, we want the result to be the sum of the two previous
values.
26. Type prev(Fib) + prev(prev(Fib))
So, the whole formula says, for the first two cases, give a value of 1; otherwise take the previous value and
add it to the value before that. This construction is typical for a recursive definition: Use caseIndex in an if
statement to define the initial condition. One fork of the if statement defines the initial values; the other
defines the recursive step.
254
Tours
A proportion can only lie between 0 and 1, so we should adjust the slider scale. We could do this by
dragging on the scale until it’s close enough, but there’s another way to control axes.
255
Fathom 2 Help
12. Make a case table for the collection, and check that you have a roughly even mix of yeses and noes.
13. Delete the case table.
14. Graph the vote attribute.
You should now have the three objects shown at right (and the inspector).
15. Choose Collection | Rerandomize several times.
Each time you rerandomize, the bars in the graph change, reflecting the results of a new sample.
16. Drag the slider’s thumb to change its value to somewhere around 0.80.
The vote is no longer close; it is a slam dunk for the proposition.
17. Move the slider’s thumb back to somewhere near 0.50 to model a close election.
256
Tours
Now that we have defined a measure, we want to collect a lot of them in a new collection (to sample our
population repeatedly). Although there’s a command that will do this, let’s use the drag-and-drop method.
22. Make a new, empty collection, putting it to the right of the existing objects.
257
Fathom 2 Help
You might want to make a summary table of the collection of samples, showing the proportion of yeses,
and run the simulation a few more times.
You could change the slider’s value, change the collect measures control to replace existing cases, and see
the results of repeated polling when the race isn’t close.
But we were investigating sample size, so, when you’re finished experimenting, let’s return to that.
We need to start the simulation over, because we’re now collecting the sample sizes.
32. Double-click the measures collection, and check Replace existing cases in the Collect Measures
panel.
33. Collect 50 measures.
Notice that the case table first empties, then fills, with two attributes this time, one for each of the
measures defined.
34. Uncheck Replace existing cases. You want to keep these measures after you change the sample
258
Tours
We now have a tool to use with the newspaper staff to show them the effects of sample size on polling
results: Smaller samples have more spread than larger, for example.
259
Fathom 2 Help
We see that most of the measurements are greater than 0, meaning that the cross-pollinated plants grew
bigger. But two of the measurements are less than 0. Darwin did not feel justified in tossing out these two
values and was faced with a very real statistical question.
Formulating a Hypothesis
Darwin’s theory—that cross-pollination produced bigger plants than self-pollination—predicts that, on
average, the difference between the two heights should be greater than 0. On the other hand, it might be
that his 15 pairs of plants have a mean difference as great as they do (21-eigths of an inch) merely by
chance. You can write out these two hypotheses in Fathom in a text object to be stored with your
document.
3. From the shelf, drag a text object into the document.
that the mean is 0, the t-statistic is simply, , where is the observed mean, s is the sample
standard deviation, and n is the number of observations.
Let’s compute this statistic for Darwin’s data using one of Fathom’s built-in statistics objects.
260
Tours
Checking Assumptions
Gosset’s work with the t-statistic relied on an assumption about the population from which
measurements would be drawn, namely, that the values in the population are normally distributed. Is this
a reasonable assumption for Darwin’s data?
Height measurements of living things, both plants and animals, are usually normally distributed, and so
are differences between heights. But we might worry, because the two negative values give a decidedly
skewed appearance to the distribution.
261
Fathom 2 Help
Fathom can help us determine qualitatively whether this amount of skew is unusual. We’ll generate
measurements randomly from a normal distribution and compare the results with the original data.
8. Make a new attribute in the collection. Call it simHeight for simulated height.
9. Select simHeight and choose Edit | Edit Formula. Enter the formula shown below.
This formula tells Fathom to generate random numbers from a normal distribution whose mean and
standard deviation are the same as in our original data. We want to compare the distribution of these
simulated heights with the distribution of the original data. We can do that directly in the dot plot that
already shows HeightDifferences.
10. Drop simHeight on the plus sign to add it to the horizontal axis.
The graph now shows the original data on top and the
simulated data on the bottom.
One set of simulated data doesn’t tell the whole story. We need to
look at a bunch.
A bit of subjectivity is called for here. Does it appear that the original distribution is very unusual, or does
it fit in with the simulated distributions?
262
Tours
12. Drag HeightDifferences from the case table to the top pane of the
test where it says “Attribute (numeric): unassigned.”
13. If the hypothesis line does not already say “is greater than,”
then select that choice from the pop-up menu.
The last paragraph of the test describes the results. If the null hypothesis were true and the experiment
were performed repeatedly, the probability of getting a value for Student’s t this great or greater would be
0.025. This is a pretty low P-value, so we can safely reject the null hypothesis and, with Darwin, pursue
the theory that cross-pollination increases a plant’s height compared with self-pollination.
14. With the test selected, choose Test | Show Test Statistic
Distribution.
The curve shows the probability density for the t-statistic with 14
degrees of freedom. The shaded area shows the portion of the area
under the curve to the right of the test statistic for Darwin’s data.
We’ve set this up as a one-tailed test; we’re only interested in the
mean difference being greater than zero. The total area under the
curve is 1, so the area of the shaded portion corresponds to the P-
value for Darwin’s experiment.
Let’s investigate how the P-value depends on the test mean, which
is currently set to 0.
15. Drag a slider from the shelf into the document.
16. Edit the name of the slider from V1 to TestMean.
17. Select the 0 in the statement of the hypothesis in the test.
Choose Edit | Edit Formula.
18. In the formula editor, enter the slider name TestMean and click OK.
Now the value of the null hypothesis mean in the test and the shaded area under the t-distribution change
to reflect the new hypothesis.
19. Drag the slider slowly and observe the changes that take place.
For what value of the slider is half the area under the curve shaded? Explain why it should be this
263
Fathom 2 Help
particular value.
The illustration below shows something similar to what you probably
have. Note that the test has been switched to “nonverbose” (choose
Test | Verbose).
Going Further
Play around with changing the data and observing the effect on the P-value. How much closer to 0 can
the experimental mean be (without changing the standard deviation) and still have a P-value greater
than 0.05? If you make the standard deviation smaller, what happens to the P-value (and why)?
Make a Test Mean object that tests the mean of simHeight instead of HeightDifferences. Notice that each
time you rerandomize, you get a new P-value. Think about what it means when the P-value is greater
than 0.05. Would you call this a “false positive” or a “false negative”? By repeatedly rerandomizing,
estimate the proportion of the time that the P-value is greater than 0.05. What practical significance
would that have in planning an experiment?
264
Tours
1. Open PetSportSurvey.ftm from the Learning Guide Starters folder in Sample Documents.
Look at the 12 cases that appear in the open collection. Simply by looking at a small number of the 325
cases in the sample, it’s impossible to make any valid predictions about what trends there might be in the
whole population.
We’ll concentrate on the question of whether there is any relationship between a student’s sex and his or
her pet preference.
2. Make a graph and drag the Sex attribute to the horizontal axis.
3. Change the graph to a ribbon chart by choosing Ribbon Chart
from the graph’s pop-up menu.
You can see from the ribbon chart the approximately equal
numbers of boys and girls in the sample.
265
Fathom 2 Help
Computing Proportions
A ribbon chart does a good job of displaying differences in proportions. But if we want to know the
computed values, we need a summary table.
266
Tours
The fractional boys and girls may seem strange, but it’s okay as long as we keep it hypothetical.
Let’s compute these numbers in the summary table.
10. Double-click the formula below the table.
11. In the formula editor, type the expression shown here and click OK.
columnTotal, rowTotal, and grandTotal are all keywords you can use only when writing a formula for a
summary table. They are in the Special list in the formula editor. You can also enter them in a formula
by double-clicking them in this list.
You should see the expected values computed in each cell.
Because this is such a common computation, Fathom has a shortcut for it.
12. Double-click the formula again and substitute the single word expected for the more complicated
expression. (This keyword is also in the Special list.) Click OK.
You should get the same results.
The term expected, or expected value, has a very general meaning in statistics. Here, in Fathom, we’re
applying it to the very particular situation of a chi-square test where the null hypothesis is that the row
and column attributes are independent.
We now want to compare the expected values with the observed values.
14. Type the formula count( ) into the formula editor and click OK.
Your summary table should look similar to the one shown at right.
Notice that the two formulas produce the same numbers for the row
and column summaries. Think about why that should be true.
267
Fathom 2 Help
There are many different statistics we could invent using the observed
and expected values. We’re going to use one for which statisticians have
figured out how to compute a distribution without having to resort to simulation—the chi-square
statistic. It’s based, as you might imagine, on the difference between the observed and expected values.
18. Show the inspector for the new collection and define a measure for
it, as shown at right. This number is the chi-square statistic.
We now have our statistic, but we do not yet know how likely it would
be to get a chi-square value this big or bigger by chance alone. The
advantage of using a well-studied statistic, however, is that Fathom can
easily compute this probability for us.
19. Define a second measure, pValue, for this collection. Give it the formula shown here.
ChiSquareCumulative is a function built in to Fathom. It takes two arguments: the first is the value of chi-
square that we have computed (the first measure we made); the second is the number of degrees of freedom
available, in this case, 1. (If you know all the row totals and column totals, degrees of freedom is the
number of cell counts you could fill in before all the rest were determined for you.) The function
computes the probability of getting that value of chi-square or less under the assumption that the two
attributes are independent. Because we’re interested in “or greater,” we subtract the function’s value from
1.
268
Tours
What are we to make of this result? We can say that if there were no difference between boys’ and girls’
pet preferences, and if we repeated the random sampling many times, we would get a result this extreme
or more extreme about 1 time in 20. For many situations, especially in the social sciences, this level of
probability is persuasive enough to say that we have probably found something. But in other situations,
especially in medical research, we would not be able to say we had found something, because the
consequences of being wrong would be too great.
269
Fathom 2 Help
It’s helpful to see where the computed chi-square statistic for this
sample lies in the distribution of chi-square values that would result
when the null hypothesis is satisfied.
24. With the test object selected, choose Test | Show Test Statistic
Distribution.
The graph you get should be similar to the one shown here. The
shaded area under the right portion of the curve corresponds to
the P-value for the observed chi-square statistic.
The null hypothesis states that there is no relationship between the two attributes Sex and Pet. What if we
were to take all the values for the attribute Sex and scramble them so that “boy” and “girl” got reassigned
randomly to each case? Any relationship that might exist between the two attributes would be wiped out
by the scrambling. Any remaining relationship would have to be due to chance alone.
25. Select the PetSportSurvey collection.
26. Choose Collection | Scramble Attribute Values.
A new collection should appear labeled Scrambled PetSportSurvey.
270
Tours
29. Make a ribbon chart for the scrambled collection, just as you did in steps 2–4.
As we scramble, we can see the variation in the relative proportions. This variation is due solely to
chance.
30. Make a new test for independence. This time, drop attributes from the scrambled collection into it.
Each time you scramble again, the chi-square statistic and the P-value are recomputed. Because you’re
simulating the conditions of the null hypothesis, the chi-square values will not be very large, and the P-
values will not be very small.
Now we want to collect many chi-square values from the scrambled collection. We will build up a
distribution of these values and see what shape it has.
31. Select the scrambled test and choose Test | Collect Results as Measures.
Fathom will scramble the scrambled collection five times, each time collecting values computed by the
test for independence and putting them in a new collection, labeled Measures from Test of
Scrambled PetSportSurvey.
32. Make a case table for the measures collection. Your screen should look similar to that shown below
(in nonverbose mode).
The two important columns in the table are chiSquareValue and pValue.
33. Make a histogram of each of the attributes, chiSquareValue and pValue.
271
Fathom 2 Help
Your histograms won’t look like much yet, because you have only collected the results of five
scrambles. You need some more.
Going Further
Consider the other two pairs of attributes possible: Sex versus Sport and Pet versus Sport. Would you
expect them to show more or less independence than Sex versus Pet? Look at the corresponding
ribbon charts. Do the observed differences in proportions look significant? Perform a chi-square test
on each of these pairs. Explain why one result is so much more significant than the other.
Though the ribbon chart is probably the easiest way to see relationships between categorical attributes,
three other displays are possible, shown below.
272
Tours
Make these three displays (using Fathom Help as needed), and learn how to interpret them. Think about
circumstances in which you might prefer one versus another.
273
Fathom 2 Help
Drag a URL: Open the file Elevator Data Web.html. Drag the
URL icon from the address bar of the Web browser into the
Fathom document.
A collection named Elevator Data Web.html appears in the
Fathom document. Open its inspector, and go to the Comments
panel. Whatever Fathom cannot decipher as data, it puts in the
collection’s comment, to help document the data.
(Extra: Copy the picture from the Web browser and paste it into
the Fathom document.)
Import from a Text File: Either drag the text file into your
document or choose File | Import | Import From File. You’ll get
a new collection, Elevator Data Text.txt, in your document. (You
could also open the text file in a word processor; select and copy
all the text; and paste into an empty collection in your Fathom
document.)
Copy and Paste from Excel: Open the Excel document Elevator Data Excel.xls. Select and copy the
entire range of filled cells. In Fathom, make an empty collection and paste in the data.
Copy and Paste from Logger Pro: If you have Logger Pro software, you can open the Elevator Data
Logger Pro file, and select and copy all the data. In Fathom, make an empty collection and paste in the
data.
3. Make a line scatter plot with Force on the vertical axis and
Time on the horizontal axis.
There about 500 points all crowded together, so it’s hard to see
the detailed structure of the graph. We can reduce the point size
to see the graph better.
4. With the graph selected, choose Object | Inspect Graph.
5. Change pointSize from 5 to 0.1.
Point size is in pixels, and this is small enough that you won’t
see any point at all.
6. Also change xAutoRescale to false. (Click on it and type.)
7. Now adjust the horizontal axis to look like the graph at right.
Have you figured out yet whether the elevator went up or down?
274
Tours
The units of acceleration are N/kg. That’s not wrong, but m/s2 would be more helpful.
275
Fathom 2 Help
Integrating Acceleration
We’ve converted the force of the elevator to acceleration.
Now, to find the distance traveled, we need to integrate
(twice). Our graph is not a smooth function, but we can
integrate over many small areas to find the velocity at any
given time. This is called numerical integration.
We’ll approximate the effect of acceleration over time by
saying that the acceleration at a given time has acted constantly
over the preceding time interval. If we magnify a small section
of the graph and add rectangles whose area is acceleration at a
given time multiplied by the time interval, as shown at right,
the area of each rectangle corresponds to an increment in the
velocity. The velocity at any given time is the sum of the areas
of the rectangles that precede it.
17. Define a new attribute, velocity, with the formula shown below.
The prev function returns the previous value of the attribute and, if you don’t specify otherwise,
returns 0 for the first case.
Notice that the units of velocity are m/s. This is a good check on the reasoning that goes into the
formula.
276
Tours
Going Further
Interpret each section of the graphs. Go for a ride in an elevator and compare that experience with
what the graphs suggest.
Investigate the sensitivity of the distance traveled to changes in restForce. If you change restForce by 0.1%,
by what percentage does the distance change?
In the tour, we assumed that the force is constant for the preceding interval of time. This clearly does
not apply to the first data point. Adjust the numerical integration to take this into account. How much
does this adjustment change the distance traveled?
277
Fathom 2 Help
internet access
a teacher account at the Fathom Surveys website. Your subscription administrator can email you
information about how to create one.
to have taken the Ethics of Classroom Surveys online tutorial
Optional: a student account that students can use to take surveys (see Add Student Users 148 for
directions)
Subsections
Making the Survey 278
A First Look at the Survey 279
Modifying the Survey 280
Taking the Survey 281
278
Tours
279
Fathom 2 Help
When you upload your survey, two changes take place at the bottom
of the Survey panel: The Upload Survey button changes to say
Survey Uploaded, and a web link, View Survey, appears above the
Download Results button.
280
Tours
You might be asking how students will find their way to this survey page. You can copy the URL from
your browser and email it to students or put it on a school site accessible to them. Or, you can write it
where they can see it and copy it into their web browser.
Another way to copy the URL is to right-click (Win) Ctrl-click (Mac)
the View Survey link and choose Copy URL from the pop-up menu.
You’ll take the survey using the student account that students will use
to take surveys.
Sometimes you, as the survey maker, want the data to come into the
collection in which you defined the survey, and other times you, as
someone other than the survey maker, want the data to come into a
new collection.
281
Fathom 2 Help
2. Make a graph to convince yourself that the data is as accessible as any other Fathom data.
Anyone can download the data into Fathom by dragging the URL of the results.
1. From the Take Survey page, click the View Results link.
2. Make a new Fathom document and position the windows on your screen so you can see both the
browser window and the Fathom document window.
3. Drag the URL from the browser into empty space in the Fathom document (see View and Download
the Results 144 ).
In the new collection the Survey panel shows the questions and has a button that can be used to
download the latest data. Unlike the Survey panel in the original collection, there is no button for
uploading the survey.
282
Tours
Explore More
Try the experiment again either by deleting all the data and starting over or by adding more data to the
existing experiment. (You’ll probably get an outlier for the first new case. Fix that by deleting its time.)
Are the two sets of data consistent?
Does your key speed depend on which row of the keyboard a given key is in? (You’ll need to figure out
how to calculate the row for each case.)
Similarly, does one hand type faster than the other? (If you’re familiar with statistical inference, this
might be an opportunity to use it!) Are some fingers faster?
What about pairs of keypresses? Does having to use one finger to type two consecutive letters (as in
the first two letters of love) slow you down?
Use your data to estimate your typing speed in words per minute. Do this both on a per character level
and as an average over the entire collection. Does the average of the instantaneous speeds equal the
overall average?
Making Meters
283
Fathom 2 Help
A meter is a Fathom object that connects to a sensor. This experiment requires a distance meter.
1. Plug the motion sensor into the LabPro and plug the LabPro into
the USB port of the computer.
2. Make a distance meter by clicking on the meter icon and dragging it
to the document.
The new meter gets the name Distance and begins reporting the distance.
Every second you should see a new case appear in the case table. If so,
you’re almost ready to gather the “real” distance data. If not, make
sure the sensor is properly connected to the computer and try again.
7. Press Turn Experiment Off to turn off the data collection.
8. Check Replace existing cases to get rid of the throwaway data.
284
Tours
You might want to change the number of cases per second in the
inspector to something more frequent—perhaps 10 cases per second
10. Make a scatter plot showing distance versus time.
11. Add a movable line to the graph (choose Graph | Add Movable
Line) and use the slope of the line to estimate your speed.
Making Meters
A meter is a Fathom object that connects to a sensor or to a slider. This experiment requires a
temperature meter.
1. Plug the temperature sensor into the LabPro and plug the LabPro
into the USB port of the computer.
2. Make a temperature meter by clicking on the meter icon and
dragging it to the document.
The new meter gets the name Temperature and begins reporting the temperature.
285
Fathom 2 Help
Every second you should see a new case appear in the case table. If so,
you’re almost ready to gather the “real” temperature data. If not, make
sure the sensor is properly connected to the computer and try again.
8. Press Turn Experiment Off to turn off the data collection.
9. Check Replace existing cases to get rid of the throwaway data.
You might want to change the number of cases per second in the
inspector to something less frequent—perhaps 5 seconds per case.
You can start analyzing the data while it is being gathered. There is no
need to stop the experiment.
11. Make a line scatter plot showing temperature versus time.
As additional data is gathered, the scatter plot updates.
Explore More
Fit a function to your data. Stop the experiment and predict the temperature of the water at some later
time. Start the experiment again and see how close (or far) you are.
Record the temperature of the air over a 24 hour period. Explain the results.
Does stirring a teaspoon of salt into glass of water change its temperature? How about stirring a bit of
baking soda into some vinegar.
286
Tours
Photogate Meters
Make sure the LabPro device is plugged into the USB port of the computer.
Sometimes you may notice that the photogate meter choice does not appear at the top of the menu even
though a photogate probe is connected. This is because the older Vernier photogates do not provide
auto-identification.
1. In a new Fathom document, make two photogate meters and an empty collection.
2. Connect each meter to the collection.
When photogate meters are created, they automatically get assigned to the digital channels on the LabPro,
the first photogate meter to the first channel, and the second meter to the second channel.
3. Block one of the photogates (using your hand or any other object) and notice which of the meters
changes.
The meters are automatically named Gate1 and Gate2 corresponding to a LabPro channel.
To measure the transit time between the two photogates, the experiment must record the time at which
first is blocked and the time at which second is blocked as a single case.
287
Fathom 2 Help
Collecting Data
Everything should be ready to collect data from the photogates.
8. Click the Turn Experiment On button in the experiment’s inspector.
9. Move the object (e.g. the ball on the ramp) through the two photogates.
10. You should see a new case appear in the experiment’s case table.
Until you press the Turn Experiment Off button, each time you move the object through the two
photogates, a new case gets generated in the experiment collection.
Data Analysis
Compute the transit time as the difference of the times at which the photogates were blocked, and the
velocity as the distance between the photogates divided by the transit time.
In a real experiment there will be some other information to record for each measurement, something on
which the transit time or velocity depends. For example, with the ball rolling down the ramp, it might be
that the angle of the ramp is varied or that the distance of the ball from the first photogate is changed.
288
Tours
The length of the object rolling through the photogate is an important measurement to compute the
velocity of the object.
Explore More
Try out the other possibilities for what to collect by changing the meters’ inspectors.
See if you can set up the two photogates to directly measure the acceleration of gravity.
7 Sample Documents
Over 300 sample documents come with Fathom. They are drawn from content areas including calculus,
painting, economics, popular culture, and biology, to name a very few. Some contain collections of
thousands of cases and many attributes, while others contain just a few cases and perhaps a single
attribute. In addition to data, some documents illustrate a mathematical or statistical concept, while others
demonstrate useful techniques in Fathom.
Under most operating systems and browsers, you should be able to click the document name to open the
document in Fathom. If you have trouble opening sample documents, see I’m Having Trouble Opening a
Fathom Document from My Browser 448 .
Finding documents relevant to a particular need can be difficult. Here are three useful techniques:
The documents are listed here in Fathom Help in the same way they are grouped in the
Sample Documents folder. Look for documents under the heading that seems most
appropriate to your need.
Use Fathom Help’s Search capability. For example, if you are looking for documents that
illustrate use of confidence intervals for proportions, try searching for “proportion.”
There is a Fathom document, Sample Documents, that lists all the documents, one to a case.
You can use graphical and filtering techniques with this collection to help find what you are
looking for (or to prove that it doesn’t exist). For example, the filter includes(Description,
“population”) might be helpful in looking for data on population growth.
Subsections
Education 290
Fathom Techniques 291
Language and the Arts 291
Learning Guide Starters 293
Mathematics 293
Science 299
Social Science 304
Statistics 315
Teaching Math with Fathom 325
Walkthrough Guide Starters 327
289
Fathom 2 Help
7.1 Education
Document name Characteristics Description
290
Sample Documents
Data from Language Arts, Religion The five longest books of the
Bible - Big books with 7800 cases, 3 attributes. bible with the number of words
Useful for Activity, Project for each chapter and verse.
291
Fathom 2 Help
A portion of George H. W.
Bush’s inaugural address
Data from Language Arts with
broken down so that each
BushInaugural 1646 cases, 1 attribute. Useful for
character is a case. Compare
Activity, Project
with other president’s speeches
or other literary documents.
Data from Language Arts with The first two paragraphs of the
Declaration 1060 cases, 1 attribute. Useful for Declaration of Independence,
Activity, Project one character per case.
Data from Language Arts with The first stanza from Goethe’s
Faust 1169 cases, 1 attribute. Useful for Faust, in German, one character
Activity, Project per case.
292
Sample Documents
7.5 Mathematics
The math sample documents are grouped into folders.
Algebra 294
Calculus 295
Geometry 296
Number Theory 296
Other Mathematics 297
293
Fathom 2 Help
Probability 297
7.5.1 Algebra
Document name Characteristics Description
Demonstrates a simple
Activity from Algebra with 900
technique for plotting surfaces.
3D Graphing cases, 4 attributes. Useful for
Could be the basis of an
Activity, Project
investigation of surfaces.
Play from Algebra with a function Shows how to draw things with
FunnyFace
plot. Useful for Project function plotting.
294
Sample Documents
7.5.2 Calculus
Document name Characteristics Description
295
Fathom 2 Help
7.5.3 Geometry
Document name Characteristics Description
Characteristics of Johnson
Data with 92 cases, 11 attributes. solids: all non-uniform convex
JohnsonSolids
Useful for Project. polyhedra with regular faces and
equal edges.
296
Sample Documents
7.5.6 Probability
Activity from Probability with 365 For each of 365 days during
Birthdays1978 cases, 3 attributes. Useful for 1978, records the number of
Activity births in the U.S. Periodic data.
297
Fathom 2 Help
298
Sample Documents
7.6 Science
The documents in this folder are grouped into the following sub-folders:
Astronomy 300
Biology 301
Chemistry and Physics 301
299
Fathom 2 Help
7.6.1 Astronomy
Data from Astronomy with 513 Displays the stars laid out in a
Planetarium cases, 21 attributes. Useful for Mercator projection. Able to
Demonstration, Project show constellations.
Data from Space Science with 696 A subset of data from Biomass
SpaceData cases, 11 attributes. Useful for Production Unit aboard the
Project International Space Station (ISS)
300
Sample Documents
7.6.2 Biology
A simulation of an ecosystem
Simulation from Biology with 200
containing predators and their
PredatorAndPrey cases, 10 attributes. Useful for
prey. It shows the population of
Activity, Project
each over time.
301
Fathom 2 Help
Temperature as a function of
time measured during heating
Data from Physics with 33 cases, 3
HeatingAndCooling and cooling of a saucepan of
attributes. Useful for Activity
water. Good function fitting
data.
Temperature as a function of
Data from Physics with 12 cases, 4 time measured during heating a
HeatingWater
attributes. Useful for Activity saucepan of water. Good
function fitting data.
Measurements of acceleration
during an elevator ride.
Data from Physics with 2000 cases,
Elevator Experiment Attributes are created for
6 attributes.Useful for Activity
velocity and displacement by
integrating the acceleration data.
Data from Physics with 953 cases, Experiment data from a solar
Solar Panel Readings
5 attributes. Useful for Project panel.
302
Sample Documents
A century’s worth of
Data from Environmental Science temperature data for 10 states
USAvgAnnualMeanTemps with 30000 cases, 8 attributes. and a number of weather
Useful for Activity, Project stations within each state.
Global warming is visible.
303
Fathom 2 Help
7.6.5 Technology
Europe 308
Latin America 309
United States 309
World 313
7.7.1 Canada
Statistics Canada provides a wealth of Canadian data from its 350 active surveys that can be used with
Fathom. E-STAT is the largest single source of numeric data on the Statistics Canada site www.statcan.ca
that is Fathom-compatible. Many datasets extracted from E-STAT are included on the Fathom 2 CD, and
on the Fathom Sample activities and Sample documents web pages. Users are encouraged to visit E-
304
Sample Documents
STAT to update and extend the data in the Fathom collections found in the sample datasets on the
Fathom CD and websites.
Statistics Canada’s E-STAT is a dynamic interactive teaching and learning tool for the education
community available at http://estat.statcan.ca. It offers an enormous warehouse of reliable and timely
statistics about Canada and its ever-changing people. E-STAT is free to educational institutions and
depository libraries, once you register. E-STAT offers a range of graphing options, complete with lessons
suitable for grade 6 and up. New users are directed to About E-STAT for further information on how to
register and on the supports provided.
E-STAT contains detailed data from the last 4 censuses and time series from over 250 different surveys
in its CANSIM database. These data can be easily transferred into Fathom. CANSIM on E-STAT is
updated once a year, generally in the summer. As of July 2004 E-STAT contains 25 million time series
grouped in over 2,000 tables.
Other datasets are listed on the StatCan Learning Resources website by teaching subject at www.statcan.
ca/english/kits/teach.htm.
Statistics Canada information is used with the permission of Statistics Canada. Users are forbidden to
copy the data and redistribute them, in an original or modified form, for commercial purposes, without
the expressed permission of Statistics Canada. Information on the availability of the wide range of data
from Statistics Canada can be obtained from Statistics Canada’s Regional Offices, its web site at http://
www.statcan.ca, and its toll-free access number 1-800-263-1136.
This folder contains one sub-folder: Canadian Census Files 307 .
Data from Economics with 103 The size of the Canadian debt in
Canadian Government Debt cases, 2 attributes. Useful for Canadian dollars from 1901 to
Activity 2001.
A wealth of demographic
Data from Demography with 46
information about each of the
Canadian Metro Areas cases, 70 attributes. Useful for
metropolitan areas of over
Activity, Project
50,000 in Canada.
305
Fathom 2 Help
306
Sample Documents
307
Fathom 2 Help
7.7.2 Europe
Data from Demography with 162 For each city, provides the
Russian City Populations cases, 4 attributes. Useful for number of males and number of
Activity, Project females.
308
Sample Documents
Data from Economics with 1104 Monthly CPI data from 1913 to
CPI by Month cases, 5 attributes. Useful for 2000. Can we detect seasonal
Activity, Project variations?
309
Fathom 2 Help
Data from Sociology with 1656 Chicago murders from 1970 and
Murders cases, 18 attributes. Useful for 1990. Very interesting for data
Project exploration.
Data from Education with 1000 SAT scores and college grade
SATGPA cases, 5 attributes. Useful for point averages. Is there any
Activity, Project relationship?
Demographic information on
Data from Demography with 254
each of Texas’ counties,
TexasCounties cases, 25 attributes. Useful for
viewable in map form through
Project
coloration.
310
Sample Documents
7.7.4.1 States
311
Fathom 2 Help
Data from History with 2000 cases, 2000 people who filled out the
MA_1910_1990 24 attributes. Useful for Activity, long form of the census in 1910
Project or 1990, half in each.
312
Sample Documents
7.7.5 World
Data from Demography with 1147 Data for 23 countries from the
CountryPopulations.23 cases, 7 attributes. Useful for United Nations Demographic
Activity, Project Yearbook, 1948-1997.
Data from Demography with 3279 data for 66 countries from the
CountryPopulations.66 cases, 7 attributes. Useful for United Nations Demographic
Activity, Project Yearbook, 1948-1997.
313
Fathom 2 Help
7.8 Sports
314
Sample Documents
7.9 Statistics
The statistics sample documents are grouped into the following sub-folders:
Correlation and Regression 315
Descriptive 317
Distributions 317
Inference 319
Modeling 322
Sampling 323
315
Fathom 2 Help
5 sliders parameterize a
Demonstration from Statistics with population a predictor and a
RegressionSimulation 200 cases, 2 attributes. Useful for response. A sample from this
Activity, Demonstration population responds to changes
in the parameters.
316
Sample Documents
7.9.2 Descriptive
7.9.3 Distributions
317
Fathom 2 Help
318
Sample Documents
7.9.4 Inference
The rest of the files in this folder are grouped into the following subfolders:
Confidence Intervals 319
Hypothesis Testing 320
319
Fathom 2 Help
320
Sample Documents
7.9.4.2.1 Independence
321
Fathom 2 Help
Builds up a sampling
distribution for a measure of
Simulation from Statistics with 200
how different a sample is from a
ChiSquare_2x3_Sim cases, 3 attributes. Useful for
population where what is being
Demonstration
measured is a total chi square
statistic.
7.9.5 Modeling
322
Sample Documents
7.9.6 Sampling
323
Fathom 2 Help
324
Sample Documents
7.10.1 Algebra 1
325
Fathom 2 Help
7.10.2 Algebra 2
Data from Algebra with 552 cases, From Teaching Math with
MaunaLoa2003
3 attributes. Useful for Activity Fathom
Data from Algebra with 510 cases, From Teaching Math with
RegionCarbon2000
9 attributes. Useful for Activity Fathom
326
Sample Documents
7.10.4 Statistics
Data from Calculus with 500 cases, From Teaching Math with
CensusTex
8 attributes. Useful for Activity Fathom
Data from Statistics with 648 cases, From Teaching Math with
PocketPennies
1 attribute. Useful for Activity Fathom
327
Fathom 2 Help
8 An Overview of Fathom
This section of Fathom Help is for those who want to understand how Fathom fits together. You won’t
learn much that’s immediately practical, but reading this section will help you get better at figuring things
out.
Fathom presents a simple interface to the beginning user: ten objects on a shelf and seven menus. You
can get started quickly and go far just by dragging attributes from a case table to a graph. But beneath
this simple surface lies enormous possibility for embellishing graphs with lines and functions, building
probability simulations, doing statistical inference via resampling techniques, constructing
parameterized mathematical models, and much more. As your interests in and capabilities with data
analysis and mathematics grow, Fathom will keep pace with you, allowing you to tap into undiscovered
features and ways of accomplishing things.
Subsections
Where Are the Data? 328
Cases 329
Attributes 330
Measures 333
Objects and Windows 333
Collections 334
Case Tables 334
Graphs 335
Summary Tables 335
Inspectors 336
Formulas and the Formula Editor 338
Sliders 339
Derived Collections 340
Statistical Objects 341
Documenting Your Work with Text Objects and Pictures 342
328
An Overview of Fathom
People Rerandomize
An open collection displays each case as a
gold ball. (You can change the placement
and appearance of the cases. See Change a person a person a person
the Appearance of Cases in a Collection
106 .)
a person a person a person
In a case table, each row represents one case, and each People
column represents one attribute.
se x height age favC olor
In a dot plot, each dot represents one case, and attributes 1 female 160 cm 17 y black
appear on axes. 2 female 171 cm 16 y red
To decide how to structure your data, ask “What does a case 3 male 172 cm 18 y orange
represent?” For example, if you are recording people’s height 4 male 163 cm 16 y orange
and sex, you’ll want to make each case a person with two
5 female 166 cm 16 y red
attributes: height and sex. Although you could record male
heights in one attribute and female heights in another 6 male 175 cm 17 y black
attribute, a case would then consist of a pair of heights;
however, this pairing would be arbitrary and would get in the
way as you analyze the data.
8.2 Cases
Without cases, there are no data. Sometimes you will encounter cases referred to as records, or even rows
(in a spreadsheet). The range of possibilities for what a case represents is vast: a planet, a car, an element,
an experiment, a set of measurements, a child, a race, a baseball player, a roll of five dice, a poker hand, a
country, a crime, the weather at a particular place and time, a purchase, a song, a family, a building.
It is essential to understand the difference between situations in which you have information about each
case and situations in which you only have summary information. Often these two kinds of data are
referred to as raw data and summary data. Fathom works well with raw data because it allows you to
explore that data and to look for relationships. Fathom does not work well if all you have is summary
data because the interesting work of deciding what to look at has already been done by the person who
created the summary.
The distinction between raw data (or microdata) and summary States
data is not necessarily clear cut. The table shows summary data
State Population Prisoners < ne w>
for each state: its population and the number of prisoners. You
could think of each person in the country as a case and realize 1 Alabama 4040587 19109
that you don’t have the raw data. But you could also consider 2 Alaska 550043 2694
each state as a case, and realize that you can do an interesting 3 Arizona 3665228 21873
analysis of the relationship between a state’s population and the 4 Arkansas 2350725 8699
number of prisoners it has.
5 California 29760021 177949
6 Colorado 3294394 13181
329
Fathom 2 Help
8.3 Attributes
Attributes describe the cases in a collection. Database mavens often call attributes fields; many
statisticians call them variables. (Fathom chooses “attribute” over “variable” to avoid confusion with the
mathematical meaning of variable.) Other, less commonly used terms are property, descriptor, and
quality.
Attributes are always of cases. In Fathom, just as you can’t have a case without a collection in which it can
live, you can’t have an attribute without a collection whose cases you want to describe. Some examples of
attributes include name (of an animal), population (of a city), speed (of a car), color (of a dress), and playingTime
(of a song).
8.3.1 Tips for Naming Attributes
Attribute names appear in formulas, and this fact places restrictions on their names. Each attribute in a
collection must have a valid, unique name. (Otherwise a formula referring to an attribute would be
ambiguous.) Valid attribute names contain only letters, underscores, and numbers; they cannot begin with
a number. No spaces or other characters are allowed. These restrictions are required for editing of
formulas to work properly.
Examples of invalid attribute names are: Person’s name (both apostrophe and space are not allowed),
IsMarried? (question mark not allowed), and 2ndYearHeight (cannot begin with a number). If you attempt
to give an attribute an invalid name, Fathom alerts you to the problem and attempts to make the name
valid by stripping out invalid characters.
To get numbers in an attribute name, put the number last or in the middle, rather than first. Instead of
1stDrop and 2ndDrop, use drop1 and drop2. When you need several words in an attribute name, separate the
words with an underscore character, or capitalize the first character in each word. Examples:
income_after_taxes or HeightBefore.
8.3.2 Values
An attribute of a given case has a value. Examples: The value of the number_of_siblings attribute for this
person is 3; the value of the diameter attribute for this wheel is 7.5 cm; the value of the color attribute for
this car is “cinnamon”. Some values are numeric, and these are either pure numbers or numbers with
units (quantities). Other values are non-numeric, or categorical.
A value can be missing. Fathom displays these values in case tables or inspectors as an empty cell with a
light yellow background. Normally, when a quantity is computed based on an attribute, the missing values
are simply not part of the computation. Likewise, a case that has a missing value for a graphed attribute
will not appear in the graph. There is no warning that some values are missing.
8.3.3 Numeric vs. Categorical Attributes
Fathom makes a fundamental distinction between attributes whose values are all numeric and attributes
that have one or more non-numeric values. The type of attribute determines how it can be graphed and
what statistics can be computed from it. For example, categorical attributes yield counts and proportions
but not means, medians, or ranges.
Most of the time, Fathom’s treatment of attributes will be what you want. But when it is not, you can
330
An Overview of Fathom
1997
2000
2002
1995
1996
1999
2001
2003
2004
Ye ar
1996 1998 2000 2002 2004 count
Ye ar To force Year to be treated as categorical, hold
Year is being treated as numeric. down the Shift key when dropping the attribute on
the graph axis.
331
Fathom 2 Help
See Manage Units in a Case Table 77 , Manage Units in the Collection’s Inspector 92 , or Units 435 .
8.3.5 Attributes with Category Sets
Categorical attributes can’t have units, but they can have
category sets assigned to them. A category set defines the valid
values that such an attribute can have. The bar chart at right, for
example, shows that the valid values of Race are “Asian”,
“Black”, “Latino”, and “White”. The bar chart leaves a place for
“Latino” even though there are no people of that race in the
collection. The bar chart also lumps all the values that are not in
the category set into a single “<invalid>” group. Using category
sets can be very helpful, and sometimes absolutely necessary, but
you can go a long way in Fathom without them.
When an attribute has a category set assigned to it, Fathom will treat it as a categorical attribute whether
or not it has any non-numeric values. The attribute for year in the U.S. census microdata, for example, is
automatically assigned a category set. That way, year can be used as a splitting attribute. Plotting year
versus income will give you, not a scatter plot, but a split dot plot.
Assigning a category set to an attribute also ensures that the data, when plotted on a graph or put in a
summary table, will be arranged by the order defined in the category set, rather than alphabetically.
332
An Overview of Fathom
8.4 Measures
You can think of a measure as a kind of attribute, but instead of
describing a single case, it describes the collection as a whole, such
as median height, maximum income, or proportion of females.
Typically, you use it to compute something from the values of the
attributes. An example is shown here. A collection consists of two
dice with an attribute face. A measure, SumFaces, has been defined
to give the total of the faces.
Simulations 156 .
Windows, such as the inspector shown below, float above the document, may be positioned outside it,
and do not scroll or print with it.
You can copy a Fathom object as a picture, which you can then paste in Fathom or other applications;
333
Fathom 2 Help
you cannot copy a window (you need a screen-capture program to get pictures of an inspector or formula
editor).
8.6 Collections
In Fathom, collections hold the data; tables and graphs help you view the data. Delete a table or graph,
and the data are still there. Delete a collection, and the data are gone.
When a collection has cases in it, its icon is a box of gold balls.
You can copy cases from one collection and paste them in another. Your
document can have any number of collections in it. measurements
When you Take a Sample 157 from a collection, the sample cases get put in another
collection. Sample of People
When you Collect Measures 159 , from a collection, the measures get put in another
collection. Similarly, when you Stack Attributes 384 , or Scramble Attribute Values
383 , you get new Derived Collections 340 .
When you Create a Survey 139 by adding questions to a collection, you get a survey
collection.
When you Create an Experiment 151 by attaching a meter to a collection, you get an
experiment collection.
334
An Overview of Fathom
8.8 Graphs
Fathom also lets you view data graphically. The graph object
supports various kinds of charts and plots (see Graphs Available
in Fathom 342 ). You tell Fathom what to graph by dragging
attribute names to an axis of the graph (see Beginner Graphing
31 ). You specify the graph type by choosing from a menu in the
335
Fathom 2 Help
Formulas appear below a summary table, and each cell of the table contains the results of those formulas
for the cases that belong to that cell. You can add any number of formulas (see Add Formula 401 ). A
menu command lets you quickly Add Basic Statistics 402 or Add Five-Number Summary 402 .
You can sort rows and columns in a summary table (Sort no data
Categories in a Summary Table 112 ), and you can extract a Drop an attribute here
summary table’s computed values into a collection, with one
1.5707963
case for each cell in the summary table, attributes that
0.13418032
describe the cell’s location, and one attribute for each formula 0.5
(Create a Collection from the Cells of a Summary Table 161 ).
S1 =
2
You can use a summary table as a calculator by adding a S2 = binomialProbability 10 ,21,0.4
formula to it to do your calculation. You can link a collection S3 = binomialCumulative 10 ,21
This summary table is being used as a calculator to calculate some values needed in the course of an
investigation.
8.10 Inspectors
When you double-click a collection, its inspector appears as a
separate window floating above the document. The inspector
shows you the names of the case attributes and allows you to
step through the collection one case at a time. There is a lot
you can do in an inspector, such as edit the values; add, delete,
and rename attributes; add, delete, and rename measures;
create and edit the collection’s comments; change how the
cases look; control the size of a sample; and turn sampling
animation on or off.
An inspector has panels, with each giving you control over a related set of characteristics of the object
you are inspecting. The tabs at the top of the inspector let you select the panel to work in. The panels that
are available in a given inspector depend on the state of the object being inspected. Following is a list of
panels that you may encounter in Fathom, with a brief description of their purpose and a link to a more
complete description of that panel’s functionality.
This panel looks at individual cases in a collection and gives you control over case
Cases Panel
attributes.
Measures Panel
This panel gives you the ability to define measures (see Define a Measure 105 ) for a
collection.
Comments Each collection has a Comments panel that you can use to describe the collection’s
Panel origin and purpose.
336
An Overview of Fathom
In addition to attributes that you define, cases in a collection always have a set of
attributes that control how they are displayed in an open collection. You can use
Display Panel
these attributes to change the position, size, image, and caption of the cases
formulaically (see Change the Appearance of Cases in a Collection 106 ).
A category set can be assigned to one or more attributes to define the valid values
for those attributes (see Attributes with Category Sets 332 ). In the Categories
Categories
panel, you can see the list of currently defined category sets and edit the categories
Panel
and their order (see Control Behavior of Categorical Attributes Using Category Sets
99 ).
Import URL When you import data from a website or from a file, the Import panel allows you
Panel to go back to that location and re-import the data.
When you import census microdata from the IPUMS site, use this panel to specify
which cases to get and which attributes the cases should have. The panel sticks with
Microdata Panel
the collection to help make it easy to go back and modify your request (see Import
U.S. Census Microdata from IPUMS 30 ).
Sampling cases from a collection creates a sample collection whose inspector has a
Sample Panel panel that gives you control over the sampling process (see Sampling Simulations
157 ).
337
Fathom 2 Help
Meter Properties A meter’s inspector has a Properties panel that gives you control over the name
Panel and the meter axis bounds
The inspector for a multiple regression model object has a Cases panel that
Multiple includes the attributes for the source collection and attributes derived by the
Regression
regression model (such as predicted values and residuals), making them available
Cases Panel
for looking at in graphs and summary tables (see Multiple Regression 203 ).
The formula editor has four main parts: the formula pane (where you actually put the formula, which tells
Fathom what you want to do), an onscreen keypad (simply click a button to put that character into the
formula pane), a list pane (from which you can double-click to input functions, attributes, and other
things into the formula), and a help pane (which provides brief help on whatever is selected in the list
pane).
The questions and links below explore further uses of formula editors.
How do I:
338
An Overview of Fathom
plot values and functions on graphs? See Add Lines and Functions to Graphs 48 .
generate random values for an attribute? See Generate Random Values 94 .
transform the values of an attribute? See Define an Attribute by Formula Using the
Collection’s Inspector 93 .
recode an attribute so that numeric codes become meaningful names? See Recode Numeric
Values to Categories 132 .
restrict the cases displayed in a graph to a subset of those in the collection? See Filter a Graph
61 .
change the computations done in a summary table? See Add or Remove a Summary Table’s
Formulas 111 .
change the heights of the bars in a bar chart to be something other than proportional to the
count? See Change Bar Charts to Reflect a Function Other Than Count 64 .
build a sampling distribution? See Collect Measures and Other Results 159 .
change the look of the gold balls in an open collection to reflect the actual data belonging to
each case? See Change the Appearance of Cases in a Collection 106 .
stop the sampling process when some condition is met, for example, a “run” of three heads?
See Make a Coin-Flip Simulation 166 .
use a slider to continuously vary the sample count in a hypothesis test, such as test mean? See
Use Sliders to Vary Summary Information in a Statistical Object 173 .
use a formula to set start conditions for an experiment? See Start Based on Conditions 155 .
8.12 Sliders
Sliders provide parameters or named numbers, in a document. You can refer to a slider value by name in
any formula in a document. The beauty of referring to a number by name is that when you decide to ask
what would happen if the number were different, you need only type in a new value or drag the slider’s
“thumb,” and you can instantly and dynamically see the effect of the change on everything in the
document that depends on it.
Sliders are commonly used as parameters to models, for example, the coefficient of a term in an equation
being fit to a set of data or the rate at which customers arrive in a queuing simulation.
Sliders have their own inspector that gives you control over quite a number of properties. Especially
powerful is the ability to compute the slider’s value with a formula. Unlike other formulas in Fathom,
slider formulas can refer to the slider itself, so that each time the formula recomputes (for example, when
the slider animates), the value changes based on the previous value. By dropping the name of a collection
on a slider, you can even compute things based on the attributes and measures in the collection.
See Work with Dynamic Parameters (Sliders) 113 and Slider Properties Panel 365 .
339
Fathom 2 Help
8.13 Meters
Meters provide an interface to external or internal sensors. You can refer to a meter value by name in any
formula in a document.
Meters are commonly used as input to experiments. You can use meters to measure the atmospheric
pressure at the time a certain liquid reaches its boiling point or the time when a car passed in front of the
window in a traffic simulation.
Meters have their own inspector, which gives you control over the name and the axis.
See Work with Meters 151 and Meter Properties Panel 366 .
The measures collection is the key to simulation and analysis. It converts measures into case attributes, so
you can record statistics (measures) about your collections. (It’s a bit tricky and is best understood after
you’ve tried it.) Suppose you have a collection with five (randomly generated) dice in it and a measure, or
total, that contains their sum. If you collect measures from that source collection, the measures
collection can record the total as a case. If you tell the measures collection to collect 100 measures, it will
instruct the dice to reroll 100 times, it will compute each total, and it will record those 100 values. See
Create Simulations 156 .
Dice Measures from Dice
face < ne w> total < ne w>
= randomPick 1 ,2,3,4,5,6 1 16
1 2 2 20
2 3 3 17
3 2 4 15
4 1 5 20
5 6 6 21
7 13
8 19
In addition to sample and measures collections, several other derived collections come in handy. A
scrambled collection reorders the values of one of the source collection’s attributes, allowing simulations of
situations in which the values of two attributes are guaranteed to be independent. A cells collection is
generated from a summary table: It has one case for each cell in the table and one attribute for each
formula. A results collection is like a measures collection, except that it treats a statistical object as the source
of its measures.
340
An Overview of Fathom
341
Fathom 2 Help
You can also paste pictures directly into Fathom. You could, Friday afternoon
for example, take a digital picture of the apparatus you used in I looked for data about track records on the internet
an experiment and put the picture in with the graphs of your and found things back to the early 1900's.
data. The most useful graphs I could make showed trends
over time.
If you write up your work in a word processing or
presentation program, copy Fathom objects as pictures and
paste them into your document. See Share Your Results:
Presenting and Printing Fathom Results 211 .
9 Reference
This section of Fathom Help provides comprehensive information about important aspects of Fathom’s
interface.
Subsections
Graphs Available in Fathom 342
Inspector Panels 353
Fathom Menus 367
Fathom Operators, Functions, and Units 413
Keyboard Shortcuts 441
You can drop attributes on the horizontal or vertical axes or in the plot area. Switch among the available
graph types by using the pop-up menu in the graph.
342
Reference
Subsections
Graphs with Numeric Attributes on One Axis 343
Graphs with Categorical Attributes on One Axis 347
Graphs with Categorical Attributes on Both Axes 350
Graphs with Numeric Attributes on Both Axes 351
See also
Dot Plot 343
Line Plot 344
Histogram 345
Ntigram 345
Box Plot 346
Percentile Plot 346
Normal Quantile Plot 347
Add Another Attribute to an Already Occupied Axis 35
343
Fathom 2 Help
The dots are stacked by default. You can unstack the dots by unchecking Graph | Stack Dots. This
aligns the dots in one line to better get an idea of sum of squares from a given value.
Split a dot plot by adding one or more numeric attributes to the same axis or by dropping a categorical
attribute on the other axis.
344
Reference
9.1.1.3 Histogram
In a histogram, cases are grouped into bins of equal width. A rectangle represents each bin, showing how
many cases are in it. Control the width of the bins by dragging a bin edge or using the graph’s inspector.
Drop a numeric attribute on an axis and choose Histogram from the graph’s pop-up menu.
You can change the vertical axis scale by choosing Frequency (the default), Relative Frequency,
Relative Percentage, or Density from the Graph | Scale menu.
Split a histogram either by adding numeric attributes to the same axis or by dropping a categorical
attribute on the other axis.
9.1.1.4 Ntigram
In an ntigram, cases are grouped into bins of equal population. The height of the corresponding
rectangle is proportional to the density of cases in the bin. You can control the population in each bin by
dragging on the bin edge or by using the graph’s inspector.
Drop a numeric attribute on an axis and choose Ntigram from the graph’s pop-up menu.
Split an ntigram either by adding numeric attributes to the same axis or by dropping a categorical attribute
on the other axis.
345
Fathom 2 Help
Drop a numeric attribute on an axis and choose Box Plot from the graph’s pop-up menu.
Split a box plot by adding more numeric attributes to the same axis or by dropping a categorical attribute
on the other axis.
Drop a numeric attribute on an axis and choose Percentile Plot from the graph’s pop-up menu.
Split a percentile plot by adding more numeric attributes to the same axis or by dropping a categorical
attribute on the other axis.
346
Reference
347
Fathom 2 Help
them.
See also
Bar Chart 348
Ribbon Chart 349
Add Another Attribute to an Already Occupied Axis 35
You can add one more categorical attribute to the axis of a bar chart to get one kind of splitting. See Add
Another Attribute to an Already Occupied Axis 35 .
Or get a different kind of splitting by dropping another categorical attribute in the plot area.
348
Reference
By default, the bar chart shows the count, but you can edit the formula to have the bars represent other
values. (For example, by changing the formula to mean(age), you would make each bar’s height represent
the mean age within each marital status category.) See Change Bar Charts to Reflect a Function Other
Than Count 64 .
You can sort the bars by choosing Graph | Sort Bars. You can rearrange the bars by dragging their
labels.
9.1.2.2 Ribbon Chart
In a ribbon chart, there is a single bar (the ribbon) broken into pieces. Each piece corresponds to one
category; its size is proportional to the population. The axis shows percentages.
Drop a categorical attribute on an axis and choose Ribbon Chart from the graph’s pop-up menu.
The developers of Fathom feel that the ribbon chart is a more useful graph than its better-known cousin,
the pie chart. People can more easily and accurately compare rectangle areas than they can the areas of pie
slices. The ribbon chart’s real power, however, comes from splitting it by dropping another categorical
349
Fathom 2 Help
attribute in the plot area: Each band then shows proportions of the new attribute’s values.
In the above graph, you can easily see that there is a much smaller proportion of men among widowed
people (most of that strip of the ribbon is made up of women).
9.1.3 Graphs with Categorical Attributes on Both Axes
These are the graphs you can make by putting a categorical attribute on one axis and another categorical
attribute on the other.
Subsections
Split Bar Chart 350
Breakdown Plot 350
350
Reference
Subsections
Scatter Plot 351
Line Scatter Plot 352
Function Plot 352
You can add more numeric attributes to one of the axes or drop any attribute in the plot area.
351
Fathom 2 Help
800
600
400
200
0
920 940 960 980
pr e s s ur e
You can add attributes to one axis or drop any attribute in the plot area.
Radiosonde Line Scatter Plot Radiosonde Line Scatter Plot
800
12.0
600
11.0 400
200
10.0
0
9.0 920 940 960 980
920 940 960 980 pr e s s ur e
pr e s s ur e te m pC
352
Reference
-8
-8 -4 0 4 8 12
Drop an attribute here. x
2. Choose Graph | Plot Function (or Plot Value) to add a function (or vertical line). Use sliders in the
formulas to manipulate the functions dynamically. See Plot Functions on a Function Plot 54 or Plot a
Function You Can Vary Dynamically (Using a Slider) 55 .
In this section, we look at the panels one by one and describe their functionality in detail.
Subsections
Cases Panel 354
Measures Panel 355
Comments Panel 356
Display Panel 356
Categories Panel 357
Import URL Panel 358
Microdata Panel 358
353
Fathom 2 Help
The Cases panel displays all the case attributes for a collection, and it displays the values for those
attributes one case at a time. Why, you ask, would you need this, given that you can use a case table to
display much more data? Here are some advantages of the Cases panel:
A Cases panel can display all the attributes compactly. By making the inspector tall and
skinny, you can use it as a kind of attribute palette.
An inspector floats above the document and doesn’t take up space in the document. Displayed
in a Cases panel, the attributes are always available for dragging, no matter what part of the
document you’re looking at.
The compact display makes it easier to rearrange attribute order than it is in a case table.
Below are things that you can do in the Cases panel and how to do them.
Move from one case to the next. Click the left or right arrows at the bottom left of the panel.
Move to the next or previous Press Ctrl (Win) Option (Mac) while clicking the left or right
selected case. arrow button.
Edit a value. Click in the cell that contains the value you want to edit.
Edit an attribute’s formula. Double-click the formula cell for the attribute.
Change the width of columns. In the column header area, drag the boundary between columns.
354
Reference
Change the height of rows. In the row header area, drag the boundary between rows.
Position the mouse pointer over the attribute name. Drag the
Use an attribute in a graph or
attribute from the inspector and drop it on the desired place in
summary table.
the graph or summary table.
Rename an Attribute in a
Double-click the attribute name and type a new name.
Collection’s Inspector.
Click in the <new> cell at the bottom of the list of attributes.
Add an Attribute to a Collection. Type the name of the new attribute. See Tips for Naming
Attributes 330 .
Delete an Attribute in an Right-click (Win) Ctrl-click (Mac) to bring up a context menu and
Inspector. choose Delete Attribute.
Drag an attribute up or down, paying attention to the
Change the order of attributes. highlighting that indicates where the attribute will be placed when
you release the mouse.
Right-click (Win) Ctrl-click (Mac) to bring up a context menu and
Add cases to the collection.
choose New Cases.
Click the Show Details button. Select the attribute. Type the
Manage Units in the Collection’s
name of the desired unit in the Units field or change the unit that
Inspector.
is there.
Create a Category Set from Click the Show Details button, and select the attribute. Choose
Existing Values. Generate from Values from the Set pop-up menu.
Create a New Category Set and Click the Show Details button, and select the attribute. Choose
Apply It to an Attribute. the name of the desired category set from the Set pop-up menu.
Remove a Category Set from an Click the Show Details button, and select the attribute. Choose
Attribute. None from the Set pop-up menu.
Following are things that you can do in the Measures panel and how to do them.
355
Fathom 2 Help
Click in the <new> cell in the Measure column and type the name of
the new measure. Measure names have the same restrictions that
Create a new measure.
attributes names have (see Tips for Naming Attributes 330 ). (Also see
Define a Measure 105 .)
Change the name of a
Double-click the measure name and edit.
measure.
Right-click (Win) Ctrl-click (Mac) the measure’s name. Choose Delete
Delete a measure.
Measure from the context menu.
Click in the corresponding cell in the Value column. Note: You cannot
Edit a measure’s value.
directly edit the value of a measure that is computed by formula.
Double-click the formula cell for the measure, and create the desired
Give a measure a formula.
formula in the formula editor.
Double-click the formula cell for the measure and use the formula
Edit an existing formula.
editor to edit the formula.
Remove the formula for a Right-click (Win) Ctrl-click (Mac) on the measure, and choose Clear
measure. Formula from the context menu.
Change column widths or
Drag boundaries in the column header or row header area.
row heights.
Make a new, empty collection. Drag a measure name from the
inspector into the new collection. The new collection will gather five
Collect measures in a new
cases whose attributes are the measures of the original collection. The
collection.
most common use of a collection’s measures is for building a sampling
distribution. See Collect Measures and Other Results 159 .
When you import data from a URL or a file, Fathom will place
text that it decides is not part of the data in the Comments
panel.
See Import Data from a Text File 81 and Import Data from the
Internet 88 .
356
Reference
Many of the sample documents make use of Fathom’s ability to make data-driven collection displays.
9.2.5 Categories Panel
The Categories panel lists category sets and the categories
within sets. There are certain situations in which it is very
important to list the valid values for an attribute, and this panel
is the place to do it.
Here are a few questions and answers about using this panel. See Control Behavior of Categorical
Attributes Using Category Sets 99 .
How do I change the column In the column or row header area, drag the boundary between
widths and row heights? columns or rows.
Right-click (Win) Ctrl-click (Mac) on the name of the category set
How can I delete a category
you wish to delete. Choose Delete Set, and the selected set will go
set?
away.
357
Fathom 2 Help
The Microdata panel has three panes: the top-left pane allows
you to choose cases and attributes to import; the top-right
pane shows the controls for the selected group; and the
bottom pane describes the current request, gives you a link to
IPUMS, tells you how many cases you have imported, and has
a button that you click to import data.
Note: You can resize the three panes by dragging the dividers
between them; and, of course, you can resize the entire
inspector window.
358
Reference
When checked, the sampling process will include animation of gold balls
from the source collection to the sample collection. All graphs and tables
will update continuously during the sampling. All this has the effect of
“Animation on” checkbox
slowing things down very much, a beneficial effect when learning about
sampling. Don’t be too quick to turn the animation off—speed is not
always a good thing.
When this option is checked, the sampling process will be done with
replacement, which means that cases may appear in the sample more than
“With replacement” once. In most situations, this is the preferred choice.
checkbox When sampling is done without replacement, once a case is chosen for the
sample, it won’t be chosen again. This means that the number of cases
sampled is limited to the number in the source collection.
If this is checked, Fathom will empty the sample collection before
starting the sampling process. For most simulations, this is the preferred
“Replace existing cases”
mode.
checkbox
If this option is not checked, Fathom will not empty the sample
collection and will continue to add each sample to the cases already there.
The idea here is that you can get Fathom to automatically collect a new
sample when anything changes in the source collection. Frequently, the
change is driven by a slider that you are dragging or animating. As the
“Collect new sample slider changes, formulas in the source collection change, and a new
when source changes” sample is drawn. Typically, measures are being collected from the sample
checkbox collection so that the information about the sample is not lost.
If there are several formulas changing, Fathom collects a new sample for
each formula change. Though this is not usually what you want, it also
usually doesn’t hurt.
Edit this value to change the number of cases that the sample collection
will sample. If the sampling is without replacement and you make this
“Cases” field
field larger than the number of cases in the source, it will automatically be
set to the number of cases after the sampling is complete.
359
Fathom 2 Help
Clicking this button shows the formula editor, ready for you to type a
true/false expression. Sampling will continue until this expression
evaluates to true. Here are some examples:
face = "heads" will continue the sampling until the value of the
face attribute for the last sample case is “heads.”
“Until condition” runlength(face) = 2 will sample until two successive values of
face are identical.
uniqueValues(coupon) = 5 will sample until there are five cases,
each of which has a different value for the attribute coupon.
Note: It is easy to get into a situation in which the sample process does
not stop! Press Esc on the computer keyboard.
Clicking this button causes a new sample to be taken. Note that there is
“Sample More Cases” also a menu command available to do this when the sample collection is
button selected. Also, if you open a sample collection, there is a Sample More
Cases button at the top right of the open collection.
360
Reference
The idea here is that you can get Fathom to automatically collect more
measures when anything changes in the source collection. Frequently, the
change is driven by a slider that you are dragging or animating. As the
“Re-collect measures slider changes, formulas in the source collection change, and more
when source changes” measures are collected. (You usually want to set number of measures to
checkbox collect to one.)
If there are several formulas changing, Fathom collects new measures
for each formula change. Though this is not usually what you want, it also
usually doesn’t hurt.
Edit this value to change the number of cases that the measures
“Measures” field
collection will collect.
Clicking this button brings up the formula editor, ready for you to type a
true/false expression. Collecting measures will continue until this
“Until condition” expression evaluates to true.
Note: It is easy to get into a situation in which the collecting measures
process does not stop! Press Esc on the computer keyboard.
Clicking this button causes more measures to be collected. Note that
“Collect More Measures” there is also a menu command available to do this when the measures
button collection is selected. Also, if you open a measures collection, there is a
Collect More Measures button at the top right of the open collection.
The Survey panel has two panes. The top pane allows you to
create questions and attributes as well as the format for the
questions; the bottom pane allows you to add instructions to
the survey, and has buttons to upload the survey, view the
survey, and download results.
When checked, the responses on the website are replaced by the cases in
Replace Data on Site
the collection. This is especially useful for teachers to clean up student
checkbox
responses and reload data that students can work with.
Clicking this button uploads the survey to the website. A dialog box
prompts you for your user name and password. The button then
Upload Survey button
becomes inactive and changes to Survey Uploaded, until you make a
change to the collection.
Type instructions here that you want the survey respondents to see. You
Instructions field
can change these as the survey evolves.
361
Fathom 2 Help
Clicking this button downloads the responses from the website into the
Download Results button collection as cases. You can click repeatedly as more people take the
survey to add cases to the collection.
View Survey link Clicking this link takes you to the survey website.
Collect By Keypress When selected, the experiment will collect one case each time a key is
button pressed. The value of the key pressed is automatically collected.
When selected, the experiment will collect one case every tenth of a
Collect By Time button second for 10 seconds by default. You can change this by using the rate
and amount fields.
Edit this field to change the number of cases the experiment will collect
Frequency field for each time unit. Use the pop-up menu to change the granularity of the
time unit from seconds to hours.
Edit this field to change the duration for which the experiment will
Duration field collect. You can specify the duration either as total number of cases
collected or total amount of time the experiment lasts.
Check this box to show the formula editor, ready for you to type a
conditional expression that evaluates to true/false. The experiment will
Trigger checkbox and
not start collecting values until this condition evaluates to true. This
condition
checking begins after you turn the experiment on (see Turn Experiment
On 362 ).
Replace existing cases Check this box to clear all cases from the experiment the next time you
checkbox run the experiment. Uncheck to keep old cases.
Click this button to start collecting cases as soon as the other criteria are
Turn Experiment On
met. You can set up the other collection parameters even when sensors
button
are not present and click this button when you plug in the sensor.
362
Reference
By default, Fathom scrambles the value in the first attribute. If there are
only two attributes, it doesn’t make any difference which one you choose.
Attribute pop-up menu
But if there are more than two attributes, you can use the pop-up menu
to choose the attribute whose values get scrambled.
Normally, changes in the collection you are scrambling do not affect the
“Collect new sample
scrambled collection until the next time you force scrambling. But if this
when source changes”
checkbox is checked, scrambling takes place each time the source
checkbox
changes.
Click this button to scramble the attribute values again. You’ll find a
“Scramble Attribute similar button at the top of the open scrambled collection. You can also
Values Again” button use a menu command in the Collection menu when the scrambled
collection is selected.
363
Fathom 2 Help
364
Reference
The Cases panel should look very familiar. It differs from the
collection’s Cases panel by having three additional attributes:
the lengths of the horizontal and vertical error bars, and point
size. You can define these attributes by formula.
The first row in the panel contains the name and value of the slider. Both
of these are editable. Of course, if you only want to change the slider’s
Slider name and value
name or type a new value, you don’t need the inspector for that; simply
do it directly in the slider.
Like an attribute or a measure, a slider can have its value determined
from a formula. Double-click the Formula field to edit the formula.
Formula There is quite a range of uses for sliders with formulas. See Define a
Slider by Formula 117 and Make a Slider Do a Random Walk 119 for
details.
365
Fathom 2 Help
When this property has no value, the slider will animate at full speed if
there is no formula and recompute the formula value once if there is a
formula.
When the property has a positive value and the slider is animated,
Fathom will move the slider or recompute the formula up to a maximum
Max_updates_per_second
of this many times per second.
Clear the field’s contents by deleting the value to remove the speed
restriction.
Note: Putting in a high value for this property will not make the slider go
faster than it would with no value.
Lower_bound and These are the bounds of the slider scale. You can also change them by
Upper_bound dragging.
This property restricts the values the slider can have. For example, a
value of 1 restricts the slider value to integers, and a value of 0.1 restricts
the slider value to a precision of one-tenth.
Restrict_to_multiples_of To return the value to blank, clear the field by deleting its contents.
Note: This restriction takes precedence over the formula, so although the
formula value might be 1.23, if the property is 0.1, the slider will have a
value of 1.2.
Normally, the axis of a slider has a scale that increases to the right. You
Reverse_scale can reverse its scale by changing the value of the property from “false” to
“true”.
The first row in the panel contains the name of the meter, which is
Meter name editable. Of course, if you only want to change the meter’s name, you
don’t need the inspector for that; simply do it directly in the meter.
Lower_bound and These are the bounds of the meter scale. You can also change them by
Upper_bound dragging as in a slider.
366
Reference
The Cases panel should look very familiar. It differs from the
collection’s Cases panel by having three additional attributes:
one for predicted values, one for residuals, and one for
standardized residuals. The formulas for these three attributes
are created by the model and are not editable.
The attributes in this panel are available for use in graphs or
summary tables.
Many command names change to reflect the selected object, such as the command for deleting the
selected object that appears in the menu as Delete Graph or Delete Slider. In this section, that
command is referred to as Delete [Object].
Other commands toggle, for example Show Text Palette and Hide Text Palette. In this section, they
are referred to as Show/Hide Text Palette.
On a Mac, Quit, Preferences, and About Fathom are in the Fathom menu. This menu is not covered
in this section, but the command locations are noted in the relevant subsections.
Subsections
File Menu 368
Edit Menu 372
367
Fathom 2 Help
Subsections
New 368
Open 368
Open Sample Document 369
Close 369
Save 369
Save As 369
Import 370
Export Collection 371
Revert Collection 371
Show Page Breaks 372
Print Preview (Win only) 372
Page Setup/Print Setup 372
Print 372
Exit/Quit 372
9.3.1.1 New
Brings up a new, blank Fathom document. A new document window appears on top of all other
windows and becomes the active window. The new document is untitled until you name it by saving it.
Shortcut: Win: Ctrl+N; Mac: z+ N
See also
Open 368
9.3.1.2 Open
Opens an existing Fathom document. When you choose Open, a dialog box appears showing files in
your current folder. (This folder depends on your platform and system settings. To open in Fathom’s
368
Reference
See also
Import From File 370
Paste 374
Copy and Paste Data from Another Program 26
See also
Exit/Quit 372
9.3.1.5 Save
Saves whatever changes have been made to the current document since the last time it was saved. If the
document is being saved for the first time, Save prompts you for a name and allows you to determine
where (in what folder) the file will be saved (see Save As 369 ).
This command is enabled only if you’ve made changes in the document since the last time you saved it.
Whenever you import, add, or change data, it’s especially important for you to save the data (by saving
the document). Because it’s easy to change data in Fathom (for example, by dragging data in a graph), you
may need to be able to use Revert Collection. This command is not available unless the data were saved
after being created or imported. Although you can Undo your way back to the original data, you also have
to undo everything, such as creating graphs and such. Revert is handier.
369
Fathom 2 Help
Subsections
Import From File 370
Import From URL 370
Import U.S. Census Data 371
9.3.1.7.1 Import From File
Imports data from a text file. Choosing this command gives you a dialog box, showing the contents of
your current folder.
1. Navigate to the folder that holds the text file you want to import.
2. Double-click the name of the file to import, or select the file and click Open.
Fathom creates a new collection that holds the data in the text file.
You should look at the data in a case table to see whether the import worked as planned. (With the
collection selected, make a new case table.)
Fathom imports text files. If your data are in some other format, export the data as a tab-delimited text
file before importing into Fathom.
See also
Import Data from a Text File 81
Copy and Paste Data from Another Program 26
Imports data from the Internet. This command is enabled when you have copied a URL.
Choosing this command invokes a dialog box into which you can type or paste a URL. (The URL must
be for the Web page that actually has the data on it.)
Dropping a URL from a browser into a Fathom document has the same effect as using this command.
(see Import Data from the Internet 88 .)
You should look at the data in a case table to see whether the import worked as planned. (With the
collection selected, make a new case table.)
You might want to check the collection’s comments; text that isn’t data might be stored here (such as
data source, explanations of attributes, and other notes).
If the import didn’t work as well as expected, you could try copying the data from the Web page and
pasting it into an empty collection.
370
Reference
See also
Copy and Paste Data from Another Program 26
Import census microdata from the Integrated Public Use Microdata Series (IPUMS) Web site.
Choosing this command creates an empty collection and shows its inspector on the Microdata panel.
Use this panel to choose the sample (years and places) and the attributes you want to see.
See also
Import U.S. Census Microdata 30
Import Historical Data from IPUMS 87
See also
Export Data to Other Programs 106
You must select a collection to make this active. (If you have entered or imported data without saving,
this command will not be available.)
This action is not undoable; a message appears to inform you of this and asks you to either go ahead or
371
Fathom 2 Help
See also
Prepare to Print a Fathom Document 214
Print Preview (Win only) 372
Share Your Results: Presenting and Printing Fathom Results 211
See also
Show Page Breaks 372
See also
Prepare to Print a Fathom Document 214
Show Page Breaks 372
Print Preview (Win only) 372
Share Your Results: Presenting and Printing Fathom Results 211
9.3.1.14 Exit/Quit
Closes all open documents and exits Fathom. You’ll be prompted to save any unsaved work in open
documents before Fathom quits.
On Mac, this command appears in the Fathom menu.
Shortcut: Win: Alt-F4; Mac: z+Q
9.3.2 Edit Menu
The Edit menu has commands for deleting, copying, and pasting and for working with text in text
objects.
Subsections
Undo 373
372
Reference
Redo 373
Cut 374
Copy 374
Copy As Picture 374
Paste 374
Delete 375
Select All 375
Edit Formula 375
Cut Formula 376
Copy Formula 376
Paste Formula 376
Clear Formula 376
Text Font 376
Text Size 376
Text Style 377
Show/Hide Text Palette 377
Preferences 377
9.3.2.1 Undo
The wording of this command changes to reflect your latest action. Choosing it undoes that action.
Fathom has unlimited Undo (and Redo); you can use it to undo your actions, one at a time, all the way
back to when you created or opened the document. Similarly, you can Redo those actions to restore the
document to the state it was in before you started undoing.
Unlimited Undo/Redo is helpful for correcting mistakes by undoing something you didn’t mean to do. It
is also helpful for diagnosing problems. If you don’t know why something happened on your screen, look
in Undo to see what your last action was. You can then Undo and Redo, toggling back and forth until
you understand what you did.
It’s also helpful for diagnosing student problems. Have the student Undo back to where they understood
what they had. Then have them Redo each step, explaining at each step what they were trying to do.
Shortcut: Win: Ctrl+Z; Mac: z+Z
9.3.2.2 Redo
This command redoes an action you have undone (its exact wording changes to reflect the action
available for redoing). If you’ve undone several steps, you can redo each of them, one by one. Redo is
available only immediately after using Undo. If you take any other action after undoing operations, you
can no longer redo the original operations.
Use Redo in combination with Undo to move backward and forward through your Fathom actions.
It’s also helpful for diagnosing student problems. Have the student Undo back to where they understood
what they had. Then have them Redo each step, explaining at each step what they were trying to do.
Shortcut: Win: Ctrl+R; Mac: z+R
See also
373
Fathom 2 Help
Undo 373
9.3.2.3 Cut
This command changes to reflect what you have selected: cases, attributes, or text. Using it removes the
selection and puts it on the clipboard (which means you can then paste it somewhere else: into another
collection or text object in the current document, in another document, or in another application).
Cut differs from Copy in that it removes the selection rather than simply putting the selection on the
clipboard.
The keyboard shortcut also works when you are editing a formula (although the menu command is not
available when the formula editor is open).
Shortcut: Win: Ctrl+X; Mac: z+X
9.3.2.4 Copy
This command changes to reflect what you have selected: cases, attributes, a collection, or text. Using it
places the selection on the clipboard (which means you can then paste it somewhere else: into another
collection or text object in the current document, in another document, or in another application).
Copy differs from Cut in that it leaves the selection in place rather than removing it. Use the related
command Copy As Picture to copy an image for pasting in Fathom or another application.
Shortcut: Win: Ctrl+C; Mac: z+C
9.3.2.5 Copy As Picture
This places an image of the selected object on the clipboard, so it can then be pasted into a blank area of
a Fathom document or a document in another application. Pictures copied in Fathom are not dynamic
and thus do not update on changes to the object or the data. (Many of the graphics in Fathom Help were
made this way.)
Shortcut: Win: Ctrl+Shift+C; Mac: z+Shift+C
9.3.2.6 Paste
The wording of this command changes to reflect what is on the clipboard and what you have selected.
(For example, if you have copied a column of data in another program and selected a collection, the
command will read Paste Attribute.)
This command puts the contents of the clipboard (whatever you have cut or copied) into the active
document. To paste cases as cases or attributes as attributes in Fathom, you must have a collection or a case
table selected. To paste a picture of a Fathom object, you cannot have any object selected in Fathom
(click in a blank area first). To paste text, paste into a text object.
Pasting cases into a collection or case table pastes all of the attributes (and any formulas defining them)
for the selected cases. If there are any same-named attributes in the collection before the paste, Fathom
appends a 1 to the attribute name.
Pasting attributes pastes the attribute name and the case values into the selected collection or case table,
but does not paste formulas. Any formula-derived values are pasted as noncomputed values.
If you have an image on the clipboard, and you select a case in an open collection, this command changes
to Paste As Case Icon. The image will replace the current case image (by default, a gold ball). (This
command won’t work if the image property has a formula; clear the formula before pasting an image as a
case icon.)
374
Reference
See also
Cut 374
Delete [Object] 379
When you have selected a(n) … The menu reads … And the command …
Collection, case table, or graph Select All Cases Selects all cases in the object
Text object Select All Text Selects all the text in the object
See also
Work with the Formula Editor 122
Plot a Line at a Given Value on a Graph 49
375
Fathom 2 Help
This command is available when a formula is selected (for instance, in a summary table or on a graph) or
when an attribute defined by a formula is selected.
Cutting a formula from an attribute does not remove the values; instead it leaves them in place as
noncomputed values.
This command is not available from the menu when a formula editor is open, but you can select the
formula and use the keyboard shortcut Ctrl+X (Win) z+X (Mac).
9.3.2.11 Copy Formula
Places a copy of the selected formula on the clipboard. The copied formula can then be pasted into a
formula editor elsewhere.
This command is not available from the menu when a formula editor is open, but you can select the
formula and use the keyboard shortcut Ctrl+C (Win) z+C (Mac).
9.3.2.12 Paste Formula
Available when an attribute is selected and a formula is on the clipboard (from a Copy or Cut command).
Choosing this command gives the selected attribute the formula from the clipboard. If the attribute
already has data, they will be replaced with the values computed by formula.
This command is not available from the menu when a formula editor is open, but you can select the
formula and use the keyboard shortcut Ctrl+V (Win) z+V (Mac).
9.3.2.13 Clear Formula
Removes the selected formula. When the selected formula is for a plotted value or function on a graph,
clearing gets rid of the plotted line or curve. When the selected formula defines an attribute’s values, the
values stay in place but are not computed: They won’t update if the data they used to depend on changes,
and they are editable.
9.3.2.14 Text Font
Provides control over what font is used in a text object. (Font control for text in other objects isn’t
available in Fathom.) If used when text is selected, the change applies to the selected text; otherwise, the
change applies to text typed after invoking the command.
This item is identical to choosing a font from the text palette.
See also
Show/Hide Text Palette 377
376
Reference
otherwise, the change applies to text typed after invoking the command.
This item is identical to changing font size via the text palette.
See also
Show/Hide Text Palette 377
Preferences 377
See also
Show/Hide Text Palette 377
See also
Text Font 376
Text Size 376
Text Style 377
9.3.2.18 Preferences
Brings up a dialog box with Fathom’s preferences with five settings.
Font Size for Text in Objects: Use the pop-up menu to choose Largest to make Fathom more
readable, for example, when projecting Fathom onto a screen for a demonstration or talk. Affects all
objects but text objects (control the size of text in text objects using the text palette or Text Size menu).
Linear Equation Form offers two choices in its pop-up menu to control how equations of lines are
formatted: y=mx+b (default) or y=a+bx.
Give Audible User Feedback allows you to turn sounds on or off.
377
Fathom 2 Help
Show Statistical Objects on the Shelf allows you to hide or show icons for the estimate, test, and
model objects in the object shelf.
New Statistical Objects are Verbose determines whether statistical objects (estimates, tests, and
models) are verbose or terse by default. (see Shorten the Output in Statistical Inference Objects 173 )
On Mac this command is in the Fathom menu.
9.3.3 Object Menu
The Object menu contains commands for actions that apply to objects in general, such as creating,
duplicating, or deleting. Actions that pertain to specific object types (for example, commands that apply
to graphs) are found in their object-specific menus (which appear when an object that uses them is
selected).
Subsections
New 378
Inspect [Object] 378
Duplicate [Object] 379
Delete [Object] 379
Add/Remove Filter 379
Hide [Object] 380
Show Hidden Object(s) 380
View In Window 380
9.3.3.1 New
Provides a submenu for creating new Fathom objects. Fathom places the new object in a blank space, if
possible, in the same area as you are viewing; if there is no blank space where you are in the document
window, Fathom places a new object below existing objects and scrolls so you can see it.
Using the commands from this menu differs from dragging objects from the shelf, in that when dragging
from the shelf, you decide where to place the object.
378
Reference
command changes, depending on the object whose inspector you will be showing.
See also
Inspectors 336
Inspector Panels 353
When text is selected, the Delete key will delete the selected text. To delete the containing object, deselect
the text by clicking the object’s frame.
Shortcut: Delete
See also
Objects and Windows 333
A filter is a Boolean expression (an expression that results in true or false values, such as sex = “female”
or age > 21) applied to an object to restrict the cases that object displays. Only cases for which the
expression is true are displayed (see Make Boolean Expressions 130 ).
379
Fathom 2 Help
To filter one object, such as a graph, apply the filter to that object; to filter all objects connected to a
collection, add the filter to the collection.
Shortcut: Win: Ctrl+F; Mac: z+F
See also
Write a Filter 121
Restrict the Cases You See by Filtering Data 96
One use for this command is to set up something hooked to a slider, but hide the slider, requiring the
user of the file to try to figure out what the slider’s value might be.
Shortcut: Win: Ctrl+Shift+H; Mac: z+Shift+H
See also
Show Hidden Object(s) 380
Delete [Object] 379
See also
Hide [Object] 380
The window can take up the entire screen space (covering the object shelf and document scroll bars, thus
enabling you to make it as big as your screen, though you may want to leave the status area of Fathom’s
window uncovered).
Click the window’s close box to get rid of the window (the object will remain).
Tip: If you are presenting findings and have a series of graphs, you can iconify the graphs. When
presenting, instead of enlarging each graph, use View In Window to quickly make it fill the screen; when
through with it, a simple click of the close bar gets rid of it.
Shortcut: Win: Alt+V; Mac: z+Option+V
See also
Share Your Results: Presenting and Printing Fathom Results 211
380
Reference
commands are available only when a collection is selected. Because case tables, and sometimes other
objects, provide ways of changing the collection they are connected to, some commands in the
Collection menu are enabled when objects connected to a collection are selected.
The Collection menu includes commands for creating and running simulations, such as building up
sampling distributions.
Subsections
Rerandomize/Sample More Cases/Scramble Again/Collect More Measures 381
New Cases 382
Prevent Changing Values in Graphs 382
Rename Collection 382
Sample Cases 382
Scramble Attribute Values 383
Collect Measures 383
Stack Attributes 384
The command
With this selected … and does this…
becomes …
Collection with at least
Rerandomizes any values determined by random
one randomly generated Rerandomize
functions in the active object.
attribute
Takes another sample from the source collection,
Sample More
Sample collection according to the settings in the Sample Cases panel
Cases
in its collection.
Randomly permutes values in the attribute chosen in
Scrambled collection Scramble Again
the Scramble panel of its inspector.
Rerandomizes or resamples the source collection, and
Collect More collects a new set of measures, according to the
Measures collection
Measures settings in the Collect Measures panel of its
inspector.
Each collection has a button, visible when the collection is open; pressing the button is equivalent to
using these commands.
381
Fathom 2 Help
See also
Create Simulations 156
This is usually the method you use to add cases to a randomly generated collection. The acts of typing or
importing data automatically create cases for you.
9.3.4.3 Prevent Changing Values in Graphs
Available when a collection, or another object connected to a collection, is selected. This command
prevents data from being dragged in graphs (which changes their values).
When the selected collection is locked, this command toggles to Enable Changing Values in Graphs.
See also
Drag Data in a Graph 39
Another way to rename a collection is to double-click its name, type a new name, and press Enter or
Return.
9.3.4.5 Sample Cases
This command creates a new collection, a sample collection, which contains ten cases randomly chosen
with replacement, from the selected (source) collection. (An animation illustrates the sampling: A blue
382
Reference
ball flies from the source to the sample collection.)Inspect the sample collection to change the default
settings (to change sample size or to sample without replacement, for example).
Using this command is equivalent to dropping a (source) collection’s name onto an empty collection.
See also
Create Simulations 156
This command creates a new, scrambled collection. By default, the first attribute of the collection has
been scrambled, that is, the values have been randomly moved around (all other attribute values are left in
place).
Scrambled Students
Change the scrambled attribute in the pop-up menu in the scrambled collection’s inspector.
This command is useful for building a simulation to test for independence of two attributes.
See also
Use Scrambling to Test for Independence 162
This command creates a new (measures) collection of five cases whose attributes are the measures from
the source collection. (This command forces the source to rerandomize, resample, or rescramble each
time a set of measures are collected.) The attributes from this collection can be dropped into graphs and
other Fathom objects in the same way as from any other collection.
This is the command you need when you want to build up a sampling distribution of a statistic (measure).
383
Fathom 2 Help
Using this command is equivalent to dropping a source collection’s measure onto an empty collection.
See also
Measures 333
Create Simulations 156
Stacked students
For example, if you have students’ heights in two attributes, one for girls’ heights, the other for boys’,
stacking will produce a collection with sex in one attribute and height in the other. (When the data do not
consist of matched pairs, then arbitrarily matching them produces cases, no meaning and prevents
analysis of them as unmatched pairs.)
384
Reference
Subsections
New Attribute 385
Rename Attribute 385
Hide Attribute(s) 385
Show Hidden Attributes 386
Number Format 386
Autofit Column Widths 387
Show/Hide Units 388
Show/Hide Formulas 388
Sort Ascending 389
Sort Descending 389
Use As Caption 389
When no attributes are selected, this command has the same effect as clicking the <new> attribute
prompt, typing a name, and pressing Enter or Return.
See also
Add Attributes in a Case Table 68
Add an Attribute to a Collection 91
See also
Rename an Attribute in a Case Table 69
Use the Shift key in conjunction with clicking to select more than one attribute.
Make hidden attributes visible using Show Hidden Attributes.
See also
Hide/Show Attributes 75
385
Fathom 2 Help
See also
Hide/Show Attributes 75
Dynamic
The default numeric format, Dynamic, is a good general-purpose format. It displays up to six significant
digits with a decimal point unless the number is too big or too small, in which case it displays scientific
notation with six significant digits.
Fixed Decimal
All numbers will have the same number of digits after the decimal point as specified by the Decimal
Places field.
Scientific Notation
A number expressed in scientific notation has one digit to the left of the decimal point, as many digits as
are specified in the Significant Digits field, and an exponent that specifies by what power of ten to
multiply.
Engineering Notation
In engineering notation, the number of digits to the left of the decimal point is either one, two, or three,
adjusted so that the power of ten will always be a multiple of three.
The chart below gives some examples, all with six digits specified in the Number Format dialog box.
386
Reference
See also
Adjust Column Width 74
387
Fathom 2 Help
The units row (when showing) appears just below the attribute names. In it, you can apply, convert, or get
rid of units for numeric attributes. If any attributes have been assigned units, the unit names appear here.
Typing a name (or recognized abbreviation) of a unit in the units row applies that unit to all values in the
attribute. Deleting the unit removes the unit from all values. Editing the unit from one to another unit in
the same dimension (of the same type, such as length or time) converts all values in the attribute from the
old to the new.
See also
Units 435
Manage Units in a Case Table 77
Manage Units in the Collection’s Inspector 92
RubberBandBallMeasurements
Group Diam e te r D_cubed <ne w >
=
Diameter
1 First 3.075 29.076
2 First 3.032 27.8732
3 First 2.951 25.6985
4 First 3.007 27.1894
5 Second 3.18 32.1574
388
Reference
See also
Work with Formulas in a Case Table 72
See also
Sort Data in a Case Table 74
Sort Data in a Graph 61
Sort Categories in a Summary Table 112
See also
Sort Data in a Case Table 74
Sort Data in a Graph 61
Sort Categories in a Summary Table 112
Captions appear under each case in an open collection and in the status bar when a case is pointed at with
the cursor.
389
Fathom 2 Help
This command is a shortcut for creating a caption formula in the collection’s Display panel.
See also
Change the Appearance of Cases in a Collection 106
Display Panel 356
Subsections
Remove X Attribute: [AttributeName] 391
Remove Y Attribute: [AttributeName] 391
Remove Legend Attribute: [AttributeName] 391
Show/Hide Axis Links 391
Rescale Graph Axes 392
390
Reference
391
Fathom 2 Help
Disconnect the axes by selecting a closed link and choosing Graph | Unlink [X/Y] Axis.
See also
Dynamically Link Axes to Each Other 43
Prevent Graph Axes from Automatically Rescaling 45
See also
Dynamically Link Axes to Each Other 43
Prevent Graph Axes from Automatically Rescaling 45
Using this command is equivalent to reselecting the graph type from the graph’s pop-up menu (for
example, re-selecting Histogram).
This command works even when the graph’s auto rescaling has been turned off.
See also
Change Axis Scales 40
Zoom In or Out in a Graph 41
Prevent Graph Axes from Automatically Rescaling 45
392
Reference
When Stack Dots is unchecked, Fathom puts cases, the same value on top of each other.
See also
Dot Plot 343
9.3.6.8 Scale
Available when a histogram is selected. Changes the nature of the scale of the vertical axis. Choose
Frequency, Relative Frequency, Relative Percentage, or Density.
See also
Histogram 345
Plot a Distribution Over a Histogram 52
393
Fathom 2 Help
When you have more than one attribute on an axis, each gets its own least-squares line.
Radiosonde Scatter Plot
12.0
11.0
10.0
9.0
0 200 400 600 800
he ight
tempC = 0.00110height + 11.1; r 2 = 0.54
dew pointC = 0.00195height + 9.2; r 2 = 0.65
See also
Make Residual Plot 396
Lock Intercept at Zero 397
Add Fitted Lines to a Graph 53
Show Squares 396
Trace on a Function or Fitted Line 56
When you have more than one attribute on an axis, each gets its own median-median line.
Radiosonde Scatter Plot Radiosonde Scatter Plot
12.5
12.0
12.0
11.5 11.0
11.0 10.0
10.5 9.0
10.0
0 200 400 600 800
9.5
he ight
9.0
dew pointC = 0.00228height + 9.1
0 200 400 600 800 tempC = 0.000913height + 11.2
he ight
dew pointC = 0.00228height + 9.1 dew pointC tempC
Re-selecting this command (unchecking it) removes the median-median line and its equation.
394
Reference
See also
Make Residual Plot 396
Lock Intercept at Zero 397
Trace on a Function or Fitted Line 56
The median-median line, like the least-squares line, is a line calculated to fit a collection of data. It differs
from the least-squares line in being more resistant to outliers and being easier to calculate. Here’s how it’s
found: Divide the data along the horizontal axis into three equally sized groups. Find the median x- and y-
values for each group (A, B, and C). Draw a line from A to C; this gives the slope of the median-median
line. Determine how far point B is from the line, and shift the line one-third of the way toward point B.
This determines the line’s location or intercept.
See also
Add Fitted Lines to a Graph 53
Dragging near an end rotates the line, dragging near the middle moves it up or down (without changing
its slope). In a univariate plot, the line can be moved left or right (up or down, if the attribute is on the
vertical axis). The cursor changes shape to suggest what dragging will do at any given moment.
See also
Remove Movable Line 395
Add a Movable Line to a Graph 395
Make Residual Plot 396
Lock Intercept at Zero 397
Show Squares 57
Show Squares 396
See also
Add Movable Line 395
Add a Movable Line to a Graph 48
395
Fathom 2 Help
Available when the selected graph has a computed line, a movable line, or a plotted value or function.
Fathom constructs a vertical line from each point to each line on the graph, then makes those segments
into sides of squares; the sum of the areas of these squares is computed and reported below the plot area.
This feature is great for demonstrating the meaning of the least-squares line (the line that minimizes the
sum of the squares of the residuals).
The squares and their total area dynamically update when the data change or a line is moved. (This can be
from moving a movable line, or changing a computed line by changing the data, or changing a function
using a slider.)
See also
Add Lines and Functions to Graphs 48
Show Squares 57
Make a Residual Plot 58
Make Residual Plot 396
Available when a scatter plot with a line or plotted function is present or is selected. Adds a new plot to
the selected graph, below the main plot, which shows each point’s vertical distance from the line or
function on the plot.
396
Reference
If you have more than one line or curve plotted on the graph, you must first specify for which line or
function you want residuals plotted by selecting the equation of the function you want used. Once the
residual plot is made, you can change its focus by selecting a different line or curve.
Data points in a residual plot can be dragged around, just like data points in the main plot. When data are
dragged, the entire graph updates dynamically. For example, if the residual plot is based on a least-squares
line, dragging a point may influence the line, and therefore the rest of the points in the residual plot will
move to reflect their residuals with respect to the (now moving) least-squares line.
See also
Make a Residual Plot 58
Plots a line at a value on the horizontal axis, determined by formula. This command shows a formula
editor into which you can put a numeric value, a slider name, or a function (such as median). A line
appears in the plot, and the formula determining it appears below the plot area (color-coded to match the
line).
In this example, three values have been plotted (one at a time): the mean and the mean plus and minus
one sample standard deviation.
See also
Plot Function 398
Show Measures of Spread in a Graph 51
397
Fathom 2 Help
Plots a function determined by formula on a bivariate graph. This command shows a formula editor into
which you can enter a function or distribution. (For example, entering only the name of the x-axis
attribute gives the line y = x.)
You can plot as many functions on a graph as you want. Clearing or cutting a function’s formula removes
that function from the graph.
See also
Plot Value 397
Plot a Function on a Graph 53
Plot a Distribution Over a Histogram 52
398
Reference
Using this command changes the order of categories in the category set (if any) applied to the attribute.
See also
Sort By Plotted Value 399
Reorder Categories in a Category Set 101
399
Fathom 2 Help
If you have more than one value plotted, select the one you want to sort by clicking its formula below the
plot area to enable the command.
See also
Sort Bars 399
Subsections
Format Value 400
9.3.7.1 Format Value
This command gives you control over how the slider displays a numeric or a date/time value. The
functionality is similar to that of the Format Attribute command from the Case Table menu (see
Format Attribute 386 ).
9.3.8 Summary Menu
The Summary menu has commands for working with summary tables. It is available when a summary
table is selected.
Subsections
Add Formula 401
Add Basic Statistics 402
Add Five-Number Summary 402
Remove Attribute 404
Sort Categories by Formula Value 404
Create Collection From Cells 404
Format Value 405
400
Reference
Boston
Total_pe rs onal_incom e
Boston 19223.189
Total_pe rs onal_incom e Female
10020
Sex
Female 19223.189 32688.097
Sex Male
Male 32688.097 22700
Column Summary 25400.853 Column Summary 25400.853
S1 = mean 16000
S1 = mean
S2 = median
Selecting a formula below the table highlights that formula’s results in the table’s cells.
Boston
Total_pe rs onal_incom e
19223.189
Female
10020
Sex
32688.097
Male
22700
Column Summary 25400.853
16000
S1 = mean
S2 = median
See also
Add Basic Statistics 402
Add Five-Number Summary 402
Sort Categories by Formula Value 404
Create Collection From Cells 404
401
Fathom 2 Help
Available when a summary table containing at least one numeric attribute is selected. Adds the following
formulas to the summary table: mean, count, sample standard deviation, missing values, and standard
error. (This command is a shortcut for repeatedly using the Add Formula command and specifying each
formula.)
Selecting a formula below the table highlights that formula’s results in the table’s cells.
Boston
TotalIncom e
19550.605
200
27517.786
1945.8013
0
S1 = mean
S2 = count
S3 = stdDev
S4 = stdError
S5 = count missing
See also
Add Formula 401
Add Five-Number Summary 402
Sort Categories by Formula Value 404
Create Collection From Cells 404
Work with Summary Tables 108
402
Reference
Boston
Total_pe rs onal_incom e
25400.853
0
4000
16000
34200
321000
S1 = mean
S2 = min
S3 = Q1
S4 = median
S5 = Q3
S6 = max
Selecting a formula below the table highlights that formula’s results in the table’s cells.
Boston
Total_pe rs onal_incom e
25400.853
0
4000
16000
34200
321000
S1 = mean
S2 = min
S3 = Q1
S4 = median
S5 = Q3
S6 = max
See also
Add Formula 401
Add Basic Statistics 402
Create Collection From Cells 404
Work with Summary Tables 108
403
Fathom 2 Help
See also
Work with Summary Tables 108
When more than one categorical attribute or more than one formula is present, the command is ghosted.
Enable it by first selecting the attribute you want to sort or the formula by which you want to sort.
See also
Sort Categories in a Summary Table 112
Work with Summary Tables 108
404
Reference
Boston
Total_pe rs onal_incom e
19223.189
243
Female
24848.422
1594.027
Sex
32688.097
206
Male
40477.813
2820.2228
Column Summary 25400.853
449
33591.478
1585.2798
S1 = mean
S2 = count
S3 = stdDev
S4 = stdError
See also
Create a Collection from the Cells of a Summary Table 161
Subsections
Verbose 405
Remove Attribute 406
Collect Results As Measures 406
9.3.9.1 Verbose
Estimates have verbose mode turned on or off according to the setting in Preferences. Whichever
setting you use, you can change the selected object’s output form using this command. Verbose mode
explains the estimate and confidence interval in common English. To better approximate other programs’
statistical output, uncheck the Verbose command.
405
Fathom 2 Help
See also
Shorten the Output in Statistical Inference Objects 173
406
Reference
You get five identical cases (the default for collecting measures). Control the collection process in the
inspector’s Collect Measures panel. By using one or more sliders in the statistical object, you can
explore statistical concepts, such as the effect of change in the data’s spread on the width of a confidence
interval.
See also
Collect Results from Statistical Objects 161
Subsections
Verbose 407
Remove Attribute 408
Collect Results As Measures 408
Show Test Statistic Distribution/Show p_hat Distribution 409
9.3.10.1 Verbose
Hypothesis tests have verbose mode turned on by default, unless you have changed this in Preferences.
Verbose mode explains the test’s results in common English. To better approximate other program’s
statistical output, uncheck Verbose.
407
Fathom 2 Help
See also
Preferences 377
408
Reference
Count: 485
Mean: 37.5196
Std dev: 22.4647 Measures from Test of Boston
Std error: 1.02007
Student's t: 2.47
DF: 484
P-value: 0.014
You get five identical cases (the default for collecting measures). Control the collection process in the
inspector’s Collect Measures panel. By using one or more sliders in the statistical object, you can
explore statistical concepts, such as the effect of change in the data’s spread on statistical results.
See also
Use Sliders to Vary Summary Information in a Statistical Object 173
Collect Measures from a Test or Estimate 210
Subsections
Verbose 410
Remove Attribute 410
Collect Results As Measures 411
409
Fathom 2 Help
9.3.11.1 Verbose
Model objects have verbose mode turned on by default, unless you change this behavior in Preferences.
Verbose mode explains the model’s results in common English. To better approximate other program’s
statistical output, uncheck the Verbose command.
Model of Cars Simple Regression Model of Cars Simple Regression
Response attribute (numeric): Hghw yMPG Response attribute (numeric): Hghw yMPG
Predictor attribute (numeric): NumCylinders Predictor attribute (numeric): NumCylinders
Sample count: 92 Sample count: 92
Equation: HghwyMPG = -2.60557
Equation of least-squares regression line: NumCylinders + 42.073
Hghw yMPG = -2.60557 Num Cylinde rs + r: -0.636174
42.073 r-squared: 0.40472
Correlation coefficient, r = -0.636174 Slope: -2.60557 +/- 0.661751
r-squared = 0.40472, indicating that 40.472% of SE Slope: 0.333095
the variation in Hghw yMPG is accounted for by Confidence level: 95 %
Num Cylinde rs.
When Num Cylinde rs = 0 , the predicted value for a future
The best estimate for the slope is -2.60557 +/- observation of Hghw yMPG is 42.0733 +/- 8.90936
0.661751 at a 95 % confidence level. (The
standard error of the slope is 0.333095.)
See also
Shorten the Output in Statistical Inference Objects 173
410
Reference
You get five identical cases (the default for collecting measures). Control the collection process in the
inspector’s Collect Measures panel. By using one or more sliders in the model, you can explore
statistical concepts.
9.3.11.4 Hide/Show Sequential Contributions Chart
Available only for multiple regression. Hides the ribbon chart shown by default in the multiple regression
model object.
Model of Airplanes Multiple Regression
See also
Multiple Regression 203
411
Fathom 2 Help
See also
Multiple Regression 203
20 40 60 80 100
20 40 60 80 100 speed range Residual
speed range Residual
Sequential Contributions
Sequential Contributions
Degrees of Sum of M e an F P
Degrees of Sum of M e an F P Sour ce Fr e e dom Squar e s Squar e Statistic V alue R2
Sour ce Fr e e dom Squar e s Squar e Statistic V alue R2
s pe e d 1 5.02001e+07 5.02001e+07 61.868 0.0000 0.7650
Re gr e s s ion 2 5.2359e+07 2.61795e+07 35.544 0.0000 0.7980
r ange 1 2.15892e+06 2.15892e+06 2.931 0.1041 0.0329
Residual 18 1.32578e+07 736547
Residual 18 1.32578e+07 736547
Total 20 6.56168e+07
Total 20 6.56168e+07
R-Squared: 0.797951
Adjusted R-Squared: 0.775501 R-Squared: 0.797951
Standard Deviation of the Error: 858.223 Adjusted R-Squared: 0.775501
Standard Deviation of the Error: 858.223
See also
Multiple Regression 203
412
Reference
Subsection
Cascade (Win only) 413
[Documents] 413
It also has commands that will take you to the Fathom Resource Center on the Web: Fathom Resource
Center takes you to the home page. Check For Updates takes you to the Web page where you can find
out what the latest version of Fathom is and download an upgrade, if necessary. Technical Support
takes you to the Web page that has a tech support form in which you can ask Key’s tech support team for
help.
On Windows computers, there is also the About Fathom splash screen in which you can see what
version of Fathom you are running. On Mac, this is found in the Fathom menu.
9.3.14 Context (Right-Click) Menus
For those of you who like to right-click, Fathom also provides context or shortcut menus. On Windows
computers, click the button on the right side of the mouse; on Macintosh computers, hold down Ctrl and
click objects to invoke these menus. This brings up a menu of items relevant to that object, collected
from many of the menu-bar menus.
For example, if you show the context menu for a graph, you’ll get many items from the Graph menu and
also items from the Collection menu, such as New Cases.
Subsections
Operators 414
413
Fathom 2 Help
9.4.1 Operators
Arithmetic operators: +, -, × , ÷, ^, !
Addition, subtraction, multiplication, division, raise to a power, and factorial. These operators obey
normal algebraic precedence, but you should use parentheses when in doubt. All of these except ! are
available from the formula editor’s keypad. You can also use the computer’s keyboard: * for
multiplication, / for division,
^ (Shift+6) to raise to a power.
414
Reference
Left and right arrows: Either on the keypad or on the computer keyboard, the arrow keys
move the caret left and right. Of course, because algebraic expressions are more complicated
than normal text, sometimes an arrow key will move the caret up or down to the preceding
or next part of the expression.
Up and down arrows: Either on the keypad or computer keyboard, the up arrow selects
more of the expression and the down arrow selects less of the selection. These, in combination
with the left and right arrows and Tab, provide a quick way to move around a complicated
expression.
Comma: Many functions use a comma to separate multiple arguments. For example, modulo
(7,3) will give 7 mod 3. You can enter the comma from the keypad, by holding down Ctrl
(Win) Option (Mac) and clicking the button, or directly from the computer keyboard.
Tab: The Tab key on the computer keyboard moves the selection to the next entry location,
for example, the next number or argument to a function. When you reach the last one, it
wraps around to the first. Shift+Tab moves in the opposite direction.
415
Fathom 2 Help
Takes a quantity (number Use this when you need to get rid of the units: scalar
scalar with units) and returns (7s) is 7. Be aware that changing units of an attribute
just the number will affect the value that scalar returns.
Returns +1 for arguments > 0; –1 for arguments <
sgn Signum function
0; 0 for arguments = 0.
Arguments for trig functions are assumed to be in radians if they don’t have units, but units may be
explicitly specified, for example, sin(90deg) is 1. The inverse functions return radians, which may be
converted to degrees: asin(1) is 1.57, and asin(1)*360deg/2ð is 90 degrees.
Regular trig functions: sin, cos, tan, sec, csc, cot. The argument is in radians and the returned value has
no units.
Inverses of regular trig functions: asin, acos, atan, asec, acsc, acot. The argument has no units and the
returned values are in radians.
Hyperbolic trig functions: sinh, cosh, tanh, sech, csch, coth. The argument is in radians and the returned
value has no units.
Inverse hyperbolic trig functions: asinh, acosh, atanh, asech, acsch, acoth. The argument has no units and
the returned values are in radians.
416
Reference
Arctangent Returns the angle in radians whose tangent is the second argument
with two divided by the first where the signs of the two arguments determine
atan2
numeric which quadrant the angle is in: atan2(0.5,-1) is –1.107 radians;
arguments atan2(-1,1)*360deg/2ð is 135 degrees.
All of these functions have an optional final argument that is a filter specifying on which cases the
function should act. For example, mean(height, Sex="male") returns the mean height of the males in the
collection.
Subsections
One Attribute Statistics Functions 417
Transformations 419
Two Attributes 420
Returns the last value in the collection for the given attribute; for example, last
last (name) would be Zelda for a collection of ducks in which the last duck’s name is
Zelda.
The median; for example, median(speed). Half the values of the attribute will be
median
above this and half will be below.
417
Fathom 2 Help
Takes two arguments, the first being the desired percentile and the second the
attribute for which the percentile will be computed. For example, percentile(50,
speed) is another way to compute the median. Or percentile(95,score) will return
percentile
the score corresponding to the 95th percentile. The first argument should be a
constant, not an attribute. If it is an attribute, the value of that attribute for the
first case will be used.
Gives the proportion of cases for which the argument is true. For example, if 12
proportion out of 24 people are over 12 years old, proportion(age > 12) will yield 0.5. When
used without an argument, this will return 1.
The value that lies at the 25th percentile; for example, the first quartile. 25% of the
Q1 values will be lower than this number and 75% will be higher. Q1(score) might give
45.
The value that lies at the 75th percentile; for example, the third quartile. 75% of
Q3 the values will be lower than this number and 25% will be higher. Q3(height) might
give 69 in.
Each of these computes the sample standard deviation according to the formula
(x-x) 2
s
sampleStdDev
n- 1
stdDev
The result is an estimate of the population standard deviation for a sample of size
N. For example, s(pressure) computes the sample standard deviation of the
attribute pressure.
Computes the square of the sample standard deviation according to the formula
(x x)2
sampleVariance n 1
. For example, sampleVariance(voltage) would compute the sample
variance of the attribute voltage.
418
Reference
Returns the standard error; for example, stdError(score). The formula used is
stdError s
n where s is the sample standard deviation and n is the number of cases.
Returns the sum of the values over all the cases. For example, sum(time)/count
sum
(isNumber(Time)) is another way to compute the mean of the attribute Time.
The number of unique values that an attribute has in the collection. For example,
uniqueValues uniqueValues(sex) will be 2 if there are only two values ("male" and "female") for
sex. (Missing values are ignored.)
Computes the variance of an attribute, that is, the square of the standard deviation,
according to the formula
(x x)2
variance n
For example, variance(before–after) computes the variance of the difference of the
two attributes before and after.
9.4.4.2 Transformations
This function helps recode numeric data to categorical by allowing you to define
histogram-like bins into which numeric values will be recoded. It takes the form bin
(a, bin, min, max) where a = attribute, bin = bin width, min = start of bin 1, and max
bin
= end. bin gives you a string (category value) for a—its “bin” as defined by the other
arguments. For example, bin(3.14, 2, 0, 10) gives “b02” because the value (3.14) is in
bin #2 in [0, 10] with bins of width 2. (The last two arguments are optional.)
The value for the next case. If this is the last case, next returns 0. For example, next
(year) returns, for each case, the value of the next year. As with prev (see below),
next next takes an optional second argument that specifies the value to be returned for
the last case. If a third argument is present, it is treated as a filter; for example, next
(height,0,Sex="f") returns the height of the next female, 0 if none.
Returns the number of population standard deviations a value is from the mean. For
popZScore example, popZScore(finalExam) computes a standard score for each value of the
attribute finalExam.
The value for the previous case. If this case is the first case, prev returns 0. For
example, prev(year) returns, for each case, the value of the previous year. A second,
optional argument allows you to specify the value that prev should take if there is no
prev previous case. For example, prev(Factor, 1) will return the previous value of Factor
for all cases except the first, for which it returns 1. If a third argument is present, it
is treated as a filter; for example, prev(numberInLine,0,Flavor="strawberry") returns the
closest previous value of numberInLine for which Flavor has the value “strawberry”.
Returns the position of the value when cases are ordered from lowest to highest.
For example, rank(Population) used as an attribute in a collection of states assigns to
rank
each state its rank according to population. Note that if there are duplicate values,
the rank will be fractional and the same for all the values. See also uniqueRank 420 .
419
Fathom 2 Help
Gives the number of identical values immediately prior to and including the current
value. For example, if flip contained {H, H, H, T, H, T, T}, this example would
runLength
return {1, 2, 3, 1, 1, 1, 2}. You could use max(runLength(flip)) to compute the
longest streak of heads or tails in a coin-flipping simulation.
Returns the number of sample standard deviations a value is from the mean. For
example, sampleZScore(height) computes a standard score for each value of the
sampleZScore attribute height. Use this function in preference to popZScore when you are working
zScore
with a sample of a population and do not know the true population standard
deviation.
Returns the unique position of a value in a list of values sorted from smallest to
largest. Each value in the list gets assigned a different rank, even if there are
uniqueRank duplicate values. For example, if attribute N contains the values {1, 2, 3, 2}, an
attribute using the expression uniqueRank(N) will have values {1, 2, 4, 3}. See also
rank.
420
Reference
Linear regression standard error of the slope. The returned value estimates the
width of the distribution of values for slope that would be obtained through
repeated sampling from the population from which the predictor and response
linRegrSESlope attributes are drawn.
Example: linRegrSESlope(Ht_Husband, Ht_Wife) returns the standard error of
the slope of the regression line that has Ht_Husband as the predictor and
Ht_Wife as the response.
Linear regression slope. Returns the slope of the least-squares regression line
linRegrSlope
with x as the independent attribute and y as the dependent attribute.
Computes the population mean of the product of the deviations of two
popCovariance random variables from their respective means. An optional filter as a third
parameter can limit the computation to certain cases.
The square of the correlation coefficient for two attributes. covariance(x, y)
rSquared represents the proportion of the variation of y that is accounted for by the
variation in x. It takes on values between 0 and 1.
Computes the sample mean of the product of the deviations of two random
sampleCovariance variables from their respective means. An optional filter as a third parameter
can limit the computation to certain cases.
421
Fathom 2 Help
randomNormal(mu, sd) A random real number pulled from a normal distribution. For
mu = the mean example, randomNormal(0, 1) gives a number from a distribution with a
sd = standard deviation mean of 0 and a standard deviation of 1.
A random nonnegative integer from a geometric distribution. Think
of the result as the number of repetitions of some event before the
result is positive given that the probability of that positive outcome is
randomGeometric(p) p.
p = the probability of Example: randomGeometric(0.5) generates the distribution of additional
a “catch.” flips of a coin necessary to get a head. Two additional parameters to
Must be between 0 and 1. the function help here. randomGeometric(p, scale, min) has possible
values min, min + scale, min + 2*scale, and so on.
So, randomGeometric(0.5, 1, 1) generates the distribution of the
number of coin flips needed to get a head.
A random real number greater than zero, pulled from a distribution
that declines exponentially (so there are more near zero, just as in the
randomExponential(mu) geometric distribution): randomExponential(5)min would be for
mu = the mean simulating times between customers when the average time between
and must be positive. customers is known to be 5 minutes.
A second, optional argument specifies the minimum value returned.
randomExponential(mu, min) will have a mean of mu + min.
dayOfMonth (date) Returns the day of the month corresponding to the given date.
Example: If date is Aug, 31, 2005, dayOfMonth( date) returns 31.
Returns the day of the week (as a number) corresponding to the given
dayOfWeek( date)
date.Example: If date is Aug, 31, 2005, dayOfWeek( date) returns 4
because August 31, 2005 is a Wednesday
Returns the short name of the day of the week corresponding to the
dayOfWeekName( date)
given date.Example: If date is Aug, 31, 2005, dayOfWeekName( date)
returns "Wed".
Returns the day of the year corresponding to the given date. Example:
dayOfYear( date)
If date is Aug, 31, 2005, dayOfYear( date) returns 243 because Aug 31
is the 243rd day of 2005.
422
Reference
Returns the short name of the month corresponding to the given date.
monthName( date)
Example: If date is Aug, 31, 2005, dayOfMonth( date) returns
"August".
now( )
Returns the current date and time as determined by the computer.
stringToDateTime( date-time- Attempts to form a date-time from the given string as though it were
string)
typed in.
ticks( ) Returns the current number of 60ths of a second since the computer
was last turned on.
423
Fathom 2 Help
0.8
0.4
0.0
-3 -2 -1 0 1 2 3
x
y = normalCumulative x
A density function returns the probability density of the distribution at a given value, x. You can plot
this function to see the shape of the distribution.
no data Function Plot
0.4
0.2
0.0
-3 -2 -1 0 1 2 3
x
y = normalDensity x
A quantile function returns the value whose cumulative probability is equal to a given value, c. You can
use this function to, for example, determine the critical value for a test statistic. You can think about it
with reference to a standardized test in which you are interested in the score on the test corresponding to
some percentile. A quantile function returns the score.
no data Function Plot
6
4
2
0
-2
-4
-3 -2 -1 0 1 2 3
x
y = normalQuantile x
See also
Plot a Distribution Over a Histogram 52
Subsections
Binomial Distribution 425
Chi-Square Distribution 425
Exponential Distribution 426
F-Distribution 427
Geometric Distribution 428
Normal Distribution 428
424
Reference
Think of this distribution in terms of coin flips. You have a certain number of coins, n, and a certain
probability of getting heads, p. This distribution deals with the probability of getting x heads. The
possibilities are 0 heads, 1 head, 2 heads, …, n heads.
Instead of going from 0 to n, you can use the min and max parameters to go from min to max in intervals
of (max – min)/n.
A plot of the cumulative distribution function when p = 0.45
The binomial cumulative distribution function computes the cumulative
binomialCumulative (x,
n, p, min, max)
probability, Pr(X ≤ x), where X is a random variable having a binomial
distribution of n choices and the probability of success equal to p.
binomialProbability (x, This probability function computes the probability that X = x, where X is
n, p, min, max) a random variable chosen from the set of possible values.
binomialQuantile (c, The binomial quantile function computes the value x, such that Pr(X ≤ x)
n, p, min, max) = c.
These functions take optional parameters that determine a minimum value and a scale so that Pr(X ≤ x),
where X = min + scale× C and C is a random variable from a chi-square distribution with df degrees of
freedom. Scale defaults to 1. Min defaults to 0.
425
Fathom 2 Help
0 1 2 3 4 5 6
x
y = chiSquareDensity x
y = chiSquareDensity x
A plot of two chi-square probability densities, one with three and the other with five degrees of freedom
chiSquareCumulative (x, The cumulative chi-square distribution function calculates the cumulative
df, scale, min) probability, Pr(X ≤ x).
chiSquareDensity (x, The chi-square density function calculates the probability density, (d/dx)Pr
df, scale, min) (X ≤ x).
chiSquareQuantile (c, The chi-square quantile function computes the value x, such that Pr(X ≤ x)
df, scale, min) = c.
The exponential functions described below take two optional parameters, scale, which defaults to 1, and
min, which defaults to 0. The standard deviation of exponentially distributed numbers equals the scale,
and the mean equals the sum of the minimum and the scale. We let E represent a random variable having
an exponential distribution.
exponentialCumulative (x, This function computes the cumulative probability, Pr(X ≤ x), where
scale, min) X = min + scale × E.
The exponential probability density function is given by
x min
exponentialDensity (x,
e s ca le
scale, min)
s ca le
426
Reference
exponentialQuantile (c, The exponential quantile function computes the value x, such that Pr
scale, min) (X ≤ x) = c. It is the inverse of exponentialCumulative.
9.4.7.4 F-Distribution
The F-distribution is used to compute p-values in a one-way analysis of variance. Under the null
hypothesis, the test statistic in this analysis (called, appropriately enough, the F-ratio) is a random variable
from an F-distribution.
The ratio referred to is the ratio of the mean square for treatments and the mean square for error. Each
of these has associated with it a number of degrees of freedom. There is a different F-distribution for
each distinct pair of degrees of freedom. These are labeled numDf and denomDf in the list that follows,
corresponding to numerator and denominator. The function plot here shows a plot of both the density
and the cumulative probability for the F-distribution, where sliders have been used to make it easy to
explore the effect of the two parameters on the shape of the distribution.
F-distribution functions also take two optional parameters: scale, which defaults to 1, and min, which
defaults to 0.
This function computes the cumulative probability, Pr(X ≤ x), where X
fCumulative(x, numDf,
denomDf, scale, min) = min + scale*F, where F is a random variable drawn from an F-
distribution with the given pair of degrees of freedom.
fDensity(x, numDf, This function calculates the probability density, (d/dx)Pr(X ≤ x).
denomDf, scale, min)
fQuantile(c, numDf, This function computes the value x, such that Pr(X ≤ x) = c.
denomDf, scale, min)
427
Fathom 2 Help
0.8
0.6
The normal distribution has the familiar 0.4
bell-shaped curve as its density function, 0.2
coming from 0.0
(( x ) / ) 2
1
2 -4 -2 0 2 4
e
2 x
where m is the mean and s is the standard y = normalDensity x
deviation of the distribution. y = normalDensity x
y = normalDensity x
Shows how the standard deviation parameter of a normal
density plot increases as the spread increases
428
Reference
Returns the number of items in the list passed as its first argument,
using the characters in its optional second argument as delimiters. The
delimiter defaults to the list separator for the current region, usually
comma.
itemCount(list, delimiters)
Examples: itemCount("a,b,c") returns 3 when comma is the regional
list separator; itemCount("JAN-1-2006", "-") returns 3;
itemCount("HTHHT", "") returns 5, because all characters are items
when there are no delimiters.
Returns the last item in the list passed as its first argument, using the
characters in its optional second argument as delimiters. The delimiter
defaults to the list separator for the current region, usually comma.
lastItem(list, delimiters) Examples: lastItem("a,b,c") returns "c" when comma is the regional
list separator; lastItem("Jane Doe", ".") returns "Doe";
lastItem("HTHHT", "") returns "T" because all characters are items
when there are no delimiters.
Returns the number of items matching its second argument in the list
passed as its first argument, using the characters in its optional third
argument as delimiters. The delimiter defaults to the list separator for
matchCount(list, match,
the current region, usually comma.
delimiters )
Examples: matchCount("2,6,3,4,3", 3) returns 2 when comma is the
regional list separator; matchCount("T;H;T;T;H", "H", ";") returns 2;
matchCount("HTHHT", "H", "") returns 3, because all characters are
items when there are no delimiters.
429
Fathom 2 Help
Returns the item at index n in the list passed as its first argument,
using the characters in its optional third argument as delimiters. The
delimiter defaults to the list separator for the current region, usually
comma.
nthItem(list, n, delimiters )
Examples: nthItem("a,b,c", 1) returns "a" when comma is the regional
list separator; nthItem("January 1, 2006", 3, ", ") returns "2006";
nthItem("HTHHT", 4, "") returns "H" because all characters are items
when there are no delimiters.
Sorts the items in the list passed as its first argument, using the
characters in its optional second argument as delimiters. The delimiter
defaults to the list separator for the current region, usually comma.
sortItems(list, delimiters) Examples: sortItems("2,16,3,4,3") returns "2,3,3,4,16" when comma is
the regional list separator; sortItems("{T;H;12;2;t;h}", "{;}") returns
"{2;12;H;h;T;t}"; sortItems("HTHHT", "") returns "HHHTT",
because all characters are items when there are no delimiters.
Takes two arguments and returns true if the second argument is a substring of the first
includes (also treated as a string): includes("the", "he") returns true; includes("dancing", "joy") returns
false; includes(1234, 23) returns true.
True when a number is in a given range: inRange(height, 60, 66) will be true for people
inRange whose height is greater than or equal to 60 and less than 66. Otherwise, assuming the
arguments are numeric, the result will be false.
True if the value is numeric: isNumber(date) will be true for the value 27 but false for the
isNumber
value June.
True if the value is a prime integer: isPrime(1234567) is false; isPrime(1234577) is true. The
isPrime argument can be as big an integer as your computer can represent, probably about 15
digits.
missing True if the value is missing (there is no value for that cell); false if there is a value present.
Returns true if the argument is odd. odd(15) returns true. odd(20) returns false. If the
odd
argument is not an integer, you will get an error.
430
Reference
Takes two arguments and returns true if the first begins with the
beginsWith(stringToLookIn, second. For example, beginsWith(LastName, "Mc") will return true
stringToFind)
for “McBride” and false for “Binker.”
Returns the ASCII code for the first character in the string
argument. For example, charToNum("Hello") is 72 because the letter
H has the ASCII code of 72. Note that even numeric arguments
charToNum(aString)
will return something for this function because the numeric result is
first converted to a string. For example, charToNum(1/2) is 48
because 48 is the ASCII code for the 0 of 0.5.
431
Fathom 2 Help
Takes two arguments and returns true if the first ends with the
endsWith(stringToLookIn, second. For example, endsWith(LastName, "er") returns true for
stringToFind)
“Binker” and false for “McBride.”
Takes three arguments, returning the position of stringToFind in
stringToFind starting from start. The first character of
stringToLookIn is numbered 1. Returns 0 if stringToFind is not
findString(stringToLookIn, found.
stringToFind, start)
The third argument need not be present and defaults to 1.
Example: findString("mathematics", "the") returns 1, but findString
("mathematics", "the", 2) returns 6.
Takes four arguments. The first is the original string. The second is
an integer > 0 specifying the starting location for the substitution.
replaceChars(aString, The third is the number of characters to be replaced, and the last is
start, numChars, substituteString) the string that is to be substituted. If numChars is 0, substituteString
is inserted.
Example: replaceChars("computer", 3, 4, "nfus") gives “confuser”.
Takes three string arguments and substitutes the third for all
replaceString(aString, occurrences of the second in the first.
stringToFind, substituteString) Example:
replaceString("12:30:45", ":", " and ") returns “12 and 30 and 45”.
432
Reference
9.4.12 Other
If there are two arguments, then the first (the value of the card) must
be between 1 and 13 and the second (the suit) must be between 1 and
4.
If there is just one argument, it must be a number between 1 and 52.
cardIcon(card, suit) This function can be used as a formula for the image attribute in the
cardIcon(cardNumber)
Display panel of a collection inspector.
Example: cardIcon(11, 2) when used for the image formula will display
the jack of diamonds in the collection.
See the Probability of a Pair 163 simulation.
With a single argument, an attribute in the collection, this returns the
index of the category corresponding to a case’s value for that attribute.
With a second argument, which should be one of the categories for the
indexOfCategory(attribute) attribute, it returns the index of that category.
indexOfCategory(attribute,
value) Examples: If the attribute sex has a category set {“M”, “F”} assigned to
it, indexOfCategory(Sex) would give 2 for a case whose value for Sex is
“F.” Similarly, indexOfCategory(Sex, "M") would give 1 and be
independent of the case.
The first two arguments are strings and refer to the name of another
collection and the name of an attribute in that collection. The third
argument is the index of a case in that collection. The result will be the
value of the attribute for that case in the collection with that name.
lookupValueByIndex Example: In a document that has a collection States,
(CollectionName, lookupValueByIndex("States", "Name", 3) would give “Arizona” as the
AttributeName, index) result.
Note: This function does not update dynamically, thus, in the example, if
the name of the third state were changed to “AZ”, this would not be
reflected as the result of the function until a recomputation were
forced.
433
Fathom 2 Help
A value equivalent to the case’s row number in a case table. No parentheses are needed.
It only makes sense when used as part of the formula of an attribute.
caseIndex
Example: modulo( caseIndex, 3) will yield {1, 2, 0, 1, 2} as values for the first five cases
in a collection.
These special values are available only for formulas in a summary table, and
columnProportion they appear in the Special heading, not in the Functions list. This is the
proportion of cases in a column of a summary table that belong in a cell.
rowProportion The proportion of cases in a row of a summary table that belong in a cell.
434
Reference
The width of bins in a histogram. This special value is only available when you are
plotting a function or value on a histogram.
binWidth
Example: binWidth*count()*normalDensity(x, mean(), s()) plots a normal curve on a
histogram with a count scale. The curve has the same area under it as has the histogram.
The number of bins in an ntigram. This special value is only available when you are
numBins
plotting a function or value on an ntigram.
9.4.17 Units
Below are the dimensions in which Fathom’s units live. All the units in a given dimension are compatible,
and Fathom can convert between them (except for the Important Constants). Some conversions are
given. Each unit lists the short form, the long form singular, the long form plural, and then other allowed
forms. You can combine units in any way you want.
Unit Dimensions
Length 435 Area 436 Volume 436
Mass 437 Time 437 Speed 438
Acceleration 438 Frequency 438 Force 438
Density 438 Pressure 438 Work 439
Power/Energy 439 Charge 439 Electrical Potential 439
Resistance 439 Current 439 Conductance 440
Capacitance 440 Magnetic Flux 440 Inductance 440
Magnetic Flux Density 440 Angle 440 Data 440
Important Constants 441 Amount 437
Length
m, meter, meters
cm, centimeter, centimeters
1 cm = 0.01 m
mm, millimeter, millimeters
1 mm = 0.001 m
km, kilometer, kilometers, click, clicks
1 km = 1000 m
nm, nanometer, nanometers
1 nm = 1 x 10^-09 m
N.M., nautical mile, nautical miles, nauticalmile, nauticalmiles, NM
1 N.M. = 1852 m
435
Fathom 2 Help
Area
ha, hectare, hectares
1 ha = 10000 m^2
acre, acres, ac
1 acre = 1/640 mi^2
barn, barns
1 barn = 1 x 10^-24 cm^2
Volume
L, liter, liters
1 L = 1000 cm^3
dL, deciLiter, deciLiters,
1 dL = 0.1L
mL, milliliter, milliliters
1 mL = 0.001 L
cc, cubic centimeter, cubic centimeters
1 cc = 1 mL
gal, gallon, gallons
1 gal = 231 in^3
qt, quart, quarts
1 qt = 1/4 gal
pt, pint, pints
1 pt = 1/2 qt
436
Reference
Mass
g, gram, grams
kg, kilogram, kilograms, kilo, kilos
1 g = 0.001 kg
mg, milligram, milligrams
1 mg = 0.001 g
MT, Metric Ton, Metric Tons
1 MT = 1000 kg
M_Earth, Earth mass, Earth masses, earth mass, earth masses, earthmass, earthmasses
1 M_Earth = 5.9742 x 10^24 kg
M_Sun, Solar mass, Solar masses, solar mass, solar masses, sunmass, sunmasses
1 M_Sun = 1.9891 x 10^30 kg
amu, atomic mass unit, atomic mass units
1 amu = 1.66053873 x 10^–27 kg
1 flOz = 1/16 pt
Amount
mole, Mole
1 mole = 6.023 x 10^23
Time
s, second, seconds, sec, secs
ns, nanosecond, nanoseconds
1ns = 0.00000000001s
µs, microsecond, microseconds,
1 µs = 0.0000001s
ms, millisecond, milliseconds
1 ms = 0.001 s
min, minute, minutes, mins
1 min = 60 s
hr, hour, hours, h
1 hr = 60 min
d, day, days
1 d = 24 hr
wk, week, weeks
1 wk = 7 d
fortnight, fortnights
1 fortnight = 14 days
yr, year, years, (Julian) year, (Julian) years, y, yrs
1 y = 365.25 days
sY, (sidereal) year, (sidereal) years
1 sY = 365.256360417 days
437
Fathom 2 Help
Speed
c, light, light, lightspeed
1 c = 2.99792458 x 10^8 m/s
mach
1 mach = 331.46 m/s
mph, mile per hour, miles per hour
1 mph = 1 mi/hr
kph, kilometer per hour, kilometers per hour
1 kph = 1 km/hr
kt, knot, knots
1 kt = 1 N.M./hr
Acceleration
gravity, gravities
1 gravity = 9.80665 m/s^2
Frequency
Hz, Hertz, hz
1 Hz = 1 s^-1
MHz, megaHertz, megahertz
1 MHz = 1,000,000 Hz
kHz, kiloHertz, kilohertz
1 kHz = 1000 Hz
Force
N, Newton, Newtons, newton, newtons
1 N = 1 kg m/s^2
dy, dyne, dynes
1 dy = 0.00001 N
kgf, kg of force, kg of force
1 kgf = 1 kg gravity
sthene
1 sthene = 1000 N
lb, pound, pounds, lbs
1 lb = 4.448222 N
oz, ounce, ounces
1 oz = 1/16 lb
Density
H2O, water
1 H2O = 1 gravity g/cm^3
Hg, Mercury
1 Hg = 13.5951 H2O
Pressure
438
Reference
Work
J, Joule, Joules, joule, joules
1 J = 1 kg m^2/s^2
erg, ergs
1 erg = 1 x 10^-7 J
BTU, BTUs
1 BTU = 1/1055.056 J
kWh, kilowatt-hour, kilowatt-hours
1 kWh = 1000 W hr
eV, electron Volt, electron Volts
1 eV = 1 electron V
Power/Energy
W, Watt, Watts, watt, watts
1 W = 1 kg m/s^3
hp, horsepower
1 hp = 550 ft lb/s
cal, cals, calorie, calories
1 cal = 4.185 J
kcal, kcals, kilocalorie, kilocalories
1kcal = 1000 cal
Charge
C, Coulomb, Coulombs, coulomb, coulombs
1C=1As
electron, electron charge, electron charges
1 electron = 1.602176462 x 10^-19 C
Electrical Potential
V, Volt, Volts, volt, volts
1 V = 1 W/A
Resistance
Ohm, Ohms, ohm, ohms
1 Ohm = 1 V/A
Current
A, Ampere, Amperes, ampere, amperes, amp amps
1 A = 1 coulomb
439
Fathom 2 Help
Conductance
S, siemens, mho
1 siemens = 1 Amp/Volt
Capacitance
F, farad, farads
1 farad = 1 coulomb/volt
uF, microfarad, microfarads
1 uF = 1,000,000 F
Magnetic Flux
Wb, weber, webers
1 weber = 1 volt second
Inductance
H, henry
1 henry = 1 weber/amp
Angle
deg, degree, degrees
rad, radian, radians
1 rad = 57.295779513 degrees
sr, steradian, steradians
1 sr = 3282.806 degrees
arcmin, arcminute, arcminutes
1 arcmin = 1/60 deg
arcsec, arcsecond, arcseconds
1 arcsec = 1/60 arcminutes
rev, revs, revolution, revolutions
1 rev = 360 degrees
Data
B, byte, bytes
kB, kilobyte, kilobytes
1 kB = 1024 B
MB, megabyte, megabytes
1 MB = 1024 kB
440
Reference
Important Constants
Planck’s constant, Planck’s constants
1 h = 6.62606876 x 10^-34 J s
G, Big G
1 G = 6.673 x 10^-11 N m^2/kg^2
Avogadro’s Number
1 mole = 6.023 x 10^23
Edit
Description Windows Mac
Shortcut Shortcut
Undo Ctrl + Z z+ Z
Redo Ctrl + R z+ R
441
Fathom 2 Help
Cut Ctrl + X z+ X
Copy Ctrl + C z+ C
Paste
Ctrl + V z+ V
442
Reference
Note: If you don’t see a new object appear after using one of these commands, the object might be hiding
behind an inspector.
Move to the previous editable attribute or value Shift + Tab Shift + Tab
Move within the active cell Left and right arrow keys Left and right arrow
keys
443
Fathom 2 Help
Move to next or previous character Right and left arrow keys Right and left arrow
keys
Move to the previous field (e.g., in a switch Shift + Tab Shift + Tab
statement)
Accept the formula and close the formula editor Enter or Return Enter or Return
Note: The formula editor, unlike Fathom in general, does not support unlimited Undo/Redo.
Preferences z+ , (comma)
Hide Fathom z+ H
Help
F1 z+ ?
10 Troubleshooting
If something on your screen has happened that you don’t understand or don’t like, the first best step is to
read what the Undo command in the Edit menu says. Choosing this command may fix your problem.
444
Troubleshooting
If you want to not only get things back the way they were but also to understand what you did, you can
use Redo in conjunction with Undo to toggle back and forth.
If undoing doesn’t help, try to find your problem in the topics within this section. If you still can’t correct
the problem, you can get technical support (see My Problem Still Isn’t Solved 454 ).
Subsections:
Problems with Data 445
Problems with Graphs 448
Problems with Sliders 451
I Can’t Drag the Slider’s Thumb 451
Problems with Inspectors 451
Problems with Formulas 452
Problems with Statistical Inference Objects 452
Problem with Fathom Surveys 453
Problems with Experiments 453
My Problem Still Isn’t Solved 454
If you still have your collection, it might have a filter on it for which no cases pass. Select the collection
and read the item for filtering in the Object menu. If it says Remove Filter rather than Add Filter, then
you have a filter on the collection for which none of the cases are true. Either remove the filter or edit it
(if you can’t see the filter, enlarge the collection by dragging its bottom corner: the filter appears below
the box of gold balls).
Here, the user meant age is greater than or equal to 21 and less than or equal to 50. Because no case satisfies
the current filter, all graphs and tables are blank.
445
Fathom 2 Help
Because the Import File command works only on text files, you need to export your data from within its
current program before trying to import.
10.1.4 A Web Page Doesn’t Import
First, make sure you are connected to the Internet and are trying to import a page that actually has the
data on it and that the data are text, not pure image, such as a picture or a PDF file. (There’s no text for
Fathom to import in these.)
If a Web page of data in text form doesn’t import by dragging its URL or using the Import from URL
command, try selecting the data, copying it, and pasting it into an empty Fathom collection. You’ll have
to copy and paste any comments separately.
Not all data will import perfectly; you might have to clean it up, either in Fathom or in another program
(such as a word processor or spreadsheet).
See Paste Data into a Collection 81 , Import Data from a Text File 81 , or Import Data from the Internet
88 .
You may have chosen attributes that are unavailable for your sample, such as Person attributes when you
are importing households, or educational attainment when you are importing cases from 1880. Check the
446
Troubleshooting
status bar comments for each attribute to see in which years it’s available.
You need to have at least one year selected for import. See Import Historical Data from IPUMS 87 .
You may have selected one or more states and one or more metro areas to import. You can choose which
states or which metro areas to import, but you can’t have selections from both. See Import Specific
Geographical Areas from IPUMS 87 .
Make sure the folder that contains the Fathom application has its Helpers folder, and that that folder has
its ImportSpecs folder, and, finally, that that folder has the IPUMS_USA_InterfaceSpec.xml file.
There may be a problem with the IPUMS Web site. Let Key Curriculum Press Tech Support know you
are having a problem by going to http://www.keypress.com/fathom/contact.htmlor try again later.
10.1.6 Fathom Treats Numeric Data as Categorical
There are two common reasons for Fathom to treat numeric data as categorical: The values of the
attribute include at least one non-numeric character that isn’t a unit (such as the letter o rather than a
zero, or an asterisk representing missing values) or the attribute has a category set applied to it.
You can use a case table with a filter not IsNumber(TroublesomeAttribute) to see those cases, non-numeric
characters. You can fix the problematic values by editing them, or you can create a new attribute using
the StringToNumber function to strip out non-numeric characters. Or you could put the attribute on the
down arrow of a summary table and scroll through it to see what the non-numeric categories are.
If all the characters are truly numeric, then the attribute might have a category set applied to it. Removing
the category set will fix the problem. (In the collection’s inspector, with Details showing, select the
attribute and choose None from the Set pop-up menu.) See Control Behavior of Categorical Attributes
Using Category Sets 99 .
If these aren’t the problem, perhaps Fathom is failing to recognize units. Make sure you have the Units
file in the right place by looking in the Fathom folder from your desktop. The Fathom application folder
(the folder that the Fathom application itself lives in) needs to contain a folder called Helpers; that folder
needs to contain a file called Units.xml. (By default, all helpers are installed correctly.) If the Helpers
folder or the Units file is missing, quit Fathom, copy the missing components from the Fathom CD,
relaunch, and try again.
10.1.7 I Assigned a Wrong Unit and Can’t Fix It
Suppose you put seconds in an attribute that is a measure of distance. If you’ve tried to simply change the
unit, the values don’t change, but they turn magenta. Fathom thinks you’re trying to convert between
incompatible units. To fix the unit, you need to first get rid of the old, and then add the correct unit.
447
Fathom 2 Help
On the Internet, files are categorized by “MIME-type” and “file extension.” Fathom documents have the
MIME type application/x-ftm and the extension .ftm.
If you use Netscape or Mozilla browsers, helper applications can be added using the Helpers tab of the
General Preferences dialog box. If you use Internet Explorer, helper applications can be added using the
File Helpers tab of the Preferences dialog box. (For other browsers, consult the browser’s
documentation.) Use your browser’s helper configuration settings to specify that Web-based data files
with the .ftm extension and/or the application/x-ftm MIME type belong to the Fathom application.
If you’re using a Macintosh, you may want to also set the downloaded file’s File Type and Creator, as
shown in the table below.
Finally, if your browser asks whether the files contain binary or text data, indicate that they are binary.
MIME Type application/x-ftm
File Extension .ftm
Macintosh File Type FDoc
Macintosh File Creator Fthm
Binary or Text? binary
Subsections:
I Want to Split a Numeric Graph, Not Make a Scatter Plot 448
I Have Frequencies: How Do I Make a Histogram? 449
I Can’t Drag Data in a Graph 449
The Text in My Graph Is Unreadable or Takes Up Too Much Space 449
The Dots Became Colored Squares 450
No Data Show in the Graph 450
Problems with Residual Plots 451
10.2.1 I Want to Split a Numeric Graph, Not Make a Scatter Plot
You get a scatter plot by putting at least one numeric attribute on each axis. Add the second attribute to
the first to split the graph by dropping the second attribute on the plus sign that appears on the axis when
you drag an attribute to the graph. See Add Another Attribute to an Already Occupied Axis 35 .
448
Troubleshooting
Age_Husband
Age_Wife
0 10 20 30 40 50 60 70
Dragging data in graphs can be disabled. To see if that’s the case, select the graph and look in the
Collection menu. If there’s a command that reads Enable Changing Values In Graphs, choose it. See
Prevent Changing Values in Graphs 382 .
10.2.4 The Text in My Graph Is Unreadable or Takes Up Too Much Space
If your attribute has a lot of categories, especially a lot of wordy categories, it’s much cleaner and easier to
read if you put the attribute on the vertical axis. You’ll probably have to make it very tall, as well. If
there’s still too much text, you can put the attribute on the down arrow of an empty summary table.
Although not the same as a graph, you can at least read all the categories by scrolling down the table. You
should consider doing a recode of the attribute to make fewer and shorter categories. (see Recode with a
Switch Statement 131 .)
449
Fathom 2 Help
200
York
aii
Portugal
Z...
District Connecticut
Vermont
Republi...
CentralColorado
Lebanon
Maine
Michigan
Mexico
Oregon
Spain
eden
Washington
Yemen ArabWisconsin
Australia andCalifornia
Alaska
Canada
England
of Columbia
Greece
Idaho
America
Florida
Indiaa
Ireland
Latvia
Louisiana
Maryland
Minnesota
Hampshire
NorthCarolina
Dakota
Other Pennsylvania
USSR/Russia
Poland
Rico
Island
Dakota
iterland
Illinois
Kansas
Massachusetts
Netherlands
Texas
West Indies
Germany
Italy
Kentucky
Jersey
AFRICA
AMERICA
Iow
Haw
New
Puerto
SwSw
New
Rhode
South
NewNew
New
North
SOUTH
count
0 10 20 30 40 50 60 70 80 90 100
Age
TotalIncom e
Perhaps the graph axes were rescaled beyond where the data lie. Choose Graph | Rescale Graph Axes
to see if that fixes the problem.
Either the graph itself, or the collection the graph is connected to, might be filtered in such a way that no
cases pass the filter. Filters are appended to the bottom of the object they filter. To see if the collection is
filtered, enlarge it. Remove a filter by selecting the object and choosing Object | Remove Filter. Edit the
filter by double-clicking it.
It is possible that you dropped a categorical attribute on a graph while pressing Ctrl (Win) Option (Mac).
This has the effect of ignoring all non-numeric characters. If the attribute has no numbers, it will be
blank.
Another possibility is that you have no data to graph. Perhaps you inadvertently deleted the collection, in
which case you should Undo until you get it back. (If the graph’s name is “No Data,” this is what has
happened.) It is also possible that you have not added cases to the collection, as when defining the
attributes by random formula. Until you tell Fathom to create cases (Collection | New Cases), you have
450
Troubleshooting
no data to display.
10.2.7 Problems with Residual Plots
When there is more than one thing plotted on the graph, the residual plot is based on the function
selected when the residual plot is made.
If you have selected a graph that has functions or computed lines on it and Make Residual Plot 396 is
disabled, you need to select the line or function on which you want the residual plot to be based. (Click
the line itself or its equation.)
If a residual plot is based on the wrong function, click the function on which you want it based, and the
residual plot will change.
0 2 4 6 8 10 12
Either clear the slider’s formula (in its inspector) or make another slider whose thumb you can drag.
10.3.2 My Slider Rounds Its Value How Can I Make It Stop?
You have a slider whose values have been restricted to multiples of some number. Remove the restriction
by clearing the Restrict_to_multiples_of field in the slider’s inspector.
Subsections
I Can’t See the Panel I Need 451
Where’s the Show Details Button? 452
451
Fathom 2 Help
Is there text next to the function name but no operator between them? If so, Fathom probably thinks the
whole thing is one word. You need to put something (multiplication, perhaps) between the items.
If nothing else works, enter the function by double-clicking it from the Functions list on the right, middle
pane of the formula editor.
10.5.2 Error Messages for Formulas
Here is a list of formula error messages and what they mean.
You have referred to a symbol that Fathom doesn’t recognize. For
#Name not recognized#
example, you may have misspelled the name of an attribute.
You have given a function an argument of the wrong type. For example,
#Type(s) incompatible#
sin("hello") gives this error.
452
Troubleshooting
Analysis of variance, test slope, and both simple and multiple linear models require raw data and cannot
be computed from summary statistics.
When you create a survey with Fathom, the collection’s Survey panel has a link that allows you to view
the survey. This is the most direct connection between Fathom and the Fathom Surveys website. See
also Log In to Your Fathom Surveys Account 148 .
You need a Fathom Surveys account (see Log In to Your Fathom Surveys Account 148 ).
10.7.3 I Uploaded a Survey, but Don't See It on the Website for Students
As a student, when you create and upload a survey, your teacher has to approve it and then publish it. If
you don't see the survey you uploaded that means your teacher has yet to approve it.
453
Fathom 2 Help
Fathom only supports Vernier sensors. If you are using a different company's sensors, Fathom will not
work with them. If you are using Vernier sensors and still experiencing problems, contact Vernier
technical support for help.
Vernier LabPro limits the user to four analog sensors and two digital sensors. Physical limitations like
device sample rates will further influence how many devices you can meaningfully work with.
11 Glossary
Glossary
attribute
A descriptor of a case, for example, the height of a person is an attribute.
bar chart
A plot of a categorical attribute in which each category has a bar whose height is determined by a formula,
count( ) by default.
bin
For a numeric attribute, a range of values. The number of cases whose attribute values lie within this
range is used to determine the height of a bar in a histogram.
bin alignment
The value on one end or the other of a bin. For a histogram, setting the bin alignment at 1.5 guarantees
that the lower bound of one bin will be 1.5.
bin width
The upper value minus the lower value of a bin.
bivariate
Involving two attributes. For example, a scatter plot is a bivariate plot because it uses an attribute on each
of its two axes.
454
Glossary
Boolean
Having a value of either true or false.
caption
The word or phrase that appears under a case icon in an open (expanded) collection. Also appears in the
status bar when the mouse pointer is over a that case’s point in a graph.
caret
The blinking indicator that shows where the next typed character will appear.
case
An individual record in a collection, for example, a person who filled out a survey, or a set of
measurements taken at a given time.
case icon
The look of a case in an open (expanded) collection. By default a gold, green, or blue ball, but
controllable by formula.
categorical
Having values drawn from a set of possibilities (categories). The race of a person and the color of a bead
are categorical attributes.
category
One of a set of possible values of an attribute, for example, “These paintings all have the category Gothic
as their value for Style.”
category set
The possible values for a categorical attribute. For example, the category set Gender consists of the values
male and female.
collect measures
To record a collection’s measures as values of a case in a measures collection. Typically, this is done
repeatedly to build up a distribution for each collected measure.
collection
The container for data in Fathom.
comments
A textual description of a collection. Can be viewed in a collection’s inspector.
cursor
Depending on context, either the position of the mouse pointer on the screen or the position of the caret
within text.
data
Generic term for the finest grain of information. Fathom is software for working with data.
distribution
The arrangement of values for an attribute with respect to their frequency. For example, the distribution
of heights for this collection of students peaks at about 170 cm.
document
Generic term for a Fathom file. A Fathom document typically contains one or more collections, graphs,
455
Fathom 2 Help
dot plot
A graph of a numeric attribute in which each case appears as a dot. In a stacked dot plot, the dots pile on
top of each other to show the shape of the attribute’s distribution.
drag
A common operation in Fathom, namely, to click, hold, and move the mouse on some object to change
its state or to link it to some other object.
export
In Fathom, to create a text file of the data contained in a collection.
Fathom
1) To measure the depth of water. 2) To understand deeply. 3) A unit of linear measure equal to 6 feet. 4)
The name of a piece of dynamic data analysis software.
file
The unit of information storage in a modern computer operating system. Fathom documents are files.
filter
In Fathom, to prevent certain cases from being displayed or used in an analysis. Any Fathom object that
displays or analyzes data can have a filter. A filter on a collection affects all objects that use that
collection’s data.
formula
In Fathom a symbolic expression consisting of attributes, functions, operators, and constants that can be
evaluated. For example, an attribute may have its values determined by a formula.
function
In Fathom a formula with one variable that takes values from a graph axis in order to show a line plot of
that function.
histogram
A plot of the values of a continuous attribute in which the heights of bars are proportional to the density
of cases whose values fall within the range of values corresponding to each bar. Fathom’s histograms
have bars of uniform width. (Compare with ntigram.)
household
One of the possible units of analysis for census microdata. Household attributes pertain to the structure
in which people live or a group of people as a whole.
hypothesis
A proposed explanation of the observed data.
image
In Fathom, the picture used as a case icon in an open (expanded) collection. By default a gold, green, or
blue ball, but can be determined formulaically or pasted onto individual cases.
import
In Fathom, to bring data into Fathom from some source that is not a Fathom document, for example,
from a text file, Web page, or census microdata server.
456
Glossary
inference
A conclusion based on evidence. In Fathom, the following objects are used for statistical inference:
interval estimates, hypothesis tests, and linear models.
inspector
In Fathom, a floating window that displays information about an object embedded in the document.
inspector pane
In Fathom, a portion of an inspector panel, for example, the left pane of the Microdata panel allows
choice of subsets of attributes to import.
inspector panel
In Fathom, one of several choices of displays of information in an inspector. A basic collection inspector
has panels for cases, measures, comments, display attributes, and categories.
interval estimate
A proposed range of reasonable values for a population parameter based on sample data. Assumes a
confidence level and typically specified as a central value plus or minus a range.
IPUMS
Integrated Public-Use Microdata Series. In Fathom, refers to the project at the University of Minnesota
Center for Population Studies from which Fathom obtains U.S. census microdata.
join
In Fathom, the process of merging two collections by using an attribute in one collection as a “key” to
cases in another (also, “table join”).
least-squares
Refers to a process of fitting a function to data in which the sum of squares of residuals is minimized.
legend
The portion of a graph that shows the key for symbols or colors used in a plot.
link axes
In Fathom, to specify that two or more numeric axis scales will have the same scales so that a change in
one will cause a change in all.
measure
In Fathom, refers to a quantity, typically computed by a formula, that applies to an entire collection.
Measures are specified in the Measures panel of a collection’s inspector.
microdata
Data about individuals, as opposed to summary data. (Each person’s sex and income in a collection of
people is raw data or microdata; proportions of females and median incomes in a collection of provinces
is summary data.) Fathom was designed to work with microdata but does less well with summary data.
model
A theoretical construct used to explain observations and predict new observations. There are many ways
to construct models in Fathom.
multiple regression
The fitting of data using more than one attribute to fit the observed values of a response attribute.
Fathom’s model object can compute a multiple linear regression, the product of which is a linear equation
457
Fathom 2 Help
multivariate
Involving multiple attributes. For example, a scatter plot with a legend attribute is a multivariate graph
because it involves three attributes.
nominal
Categorical, having named values.
ntigram
A graphical representation of the values of an attribute in which each of several bars arranged along a
numeric axis has equal area and that area is proportional to the count of cases whose values for the
attribute lie within the range of the bin corresponding to each bar.
numeric
Having a value that is a number (sometimes called “continuous”). In contrast to categorical attributes.
object
In Fathom, selectable entities within a Fathom document. A graph and a case table are objects, but an
inspector window is not because it floats above the document rather than living within it.
operator
In a Fathom, formula a symbol that indicates something to be done to one or more values. For example,
the division operator divides one value by another and returns the result.
parameter
A value that determines which of a family of functions or models is under consideration at the moment.
In Fathom, sliders are used as parameters.
percentile
Given a value for an attribute, the percent of values for that attribute that are less than or equal to the
given value.
plotted function
In Fathom, a line or curve on a plot determined by a formula using one axis as its domain for a dummy
variable in the formula.
plotted value
A line on a plot perpendicular to an axis, where the position of the line along the axis is determined by a
formula.
properties
In Fathom properties are attributes of objects other than collections. Graphs and sliders have properties
whose values determine their appearance and behavior.
quantity
A number with a unit. Attribute values in Fathom can be either plain numbers or quantities.
raw data
The data as it was originally gathered, as opposed to summary data.
recode
To take the values of one attribute and use them to determine (usually formulaically) the values of
458
Glossary
another attribute. Recoding numeric choices to their corresponding textual meaning can greatly increase
one’s ability to understand data.
redo
In computer software, refers to the act of making something that was undone happen again.
residual
The observed value minus the predicted value.
revert collection
In Fathom, to return the values of a collection’s attributes to their last saved values without changing
other objects in a document.
ribbon chart
A plot in which a rectangle representing all the cases is divided into sub-portions using one or two
categorical attributes, such that the area of each sub-portion is proportional to the count of cases having
certain values of the categorical attributes.
sample
To draw cases from a population. In Fathom, sampling is nearly always random.
scatter plot
A graphical representation in which numeric attributes placed on each axis determine the coordinates of
points representing cases in the collection.
scramble
In Fathom, to take one attribute in a collection and reassign its values randomly to that attribute.
Scrambling is typically used to simulate a situation in which two attributes are independent of each other.
shelf
In Fathom, the place at the top of the application (Win) or document (Mac) window that contains iconic
representations of the objects that can be placed in the document.
simulation
To represent some process (for example, rolling a die) through some other means, usually for the
purpose of discovering properties of that process. Fathom provides many tools for constructing
simulations.
slider
In Fathom, an object representing a named number or quantity such that the name can be used in
formulas to represent the slider’s value. The slider’s value can be determined by the position of its
“thumb,” by typing a new value, or by formula.
stack attributes
In Fathom, the process of reorganizing the data so that there are two attributes, named “group” and
“value” by default, where the values of the group attribute are the names of the attributes. The process
can be visualized as taking each attribute of the collection and stacking it under its predecessor.
statistic
A quantity computed from data.
status bar
A pane found at the bottom of an application (Windows) or document (Mac) window that displays some
459
Fathom 2 Help
string
Textual information. An attribute treated as categorical typically has string values rather than numeric
values.
summary data
Data that summarizes raw data. Summary data typically consists of many fewer values than its raw data,
and contains less “information” than the raw data. Proportions and means are summary data; values
about individuals are raw data, from which summary statistics are computed.
table join
The operation of merging two collections by dragging an attribute from one collection onto an attribute
name appearing in a case table. The result is that information from the first collection is brought into the
second using the attribute dropped upon as a key to the dragged attribute.
test
In Fathom, “test” typically refers to a hypothesis test object used to determine whether certain
observations can be explained by chance or not.
test statistic
A computed quantity whose distribution, when the null hypothesis is true, is known or can be
determined. For a given inference, the test statistic is used to determine how likely it is that the data can
be explained by the null hypothesis.
undo
In software applications, to reverse the most recent action. Fathom’s undo capabilities allow you to step a
document back to the state at which it was last opened (or created).
unit
The standard of measurement in a physical dimension to be considered as 1; for example, an attribute
value of 3.4 Joules has Joules as its unit.
univariate
Pertaining to something determined by one variable or attribute. In Fathom a dot plot with a single
attribute is a univariate plot. The mean and median are univariate measures.
URL
Uniform Resource Locator. The address of a given Web page. Fathom’s Import From URL command
takes a URL and attempts to convert the indicated page into a collection of data.
value
The number, quantity, or string corresponding to a given attribute for each case. An attribute need not
have a value for a given case, in which situation the value is said not to exist.
window
In Fathom, documents and inspectors appear in windows. An object in a document, such as a graph,
however, is not a separate window because it is constrained to stay within the document window.
zoom
To expand or contract a numeric scale on a graph. To “zoom in” is to expand the scale so that points
appear farther apart.
460
Index 461
Index -P-
Paste 26
-A- Probability 163
attributes 330
-R-
-B- recode 131, 132, 134
Redo 32
boolean 130 Reference 342
running sum 131
-C- -S-
Case Tables 334
Cases 329 Sliders 339
Collections 334 Statistical Objects 341
condition 130 Summary Tables 335
Copy 26 switch statement 131
-D- -T-
data 24, 25, 328 Troubleshooting 444
-F- -U-
faq 135 Undo 32
formulas 135 units 134
-G- -V-
Glossary 454 values 330
Graphing 31
Graphs 335
-Z-
-I- zoom 41
if statement 130
Inspectors 336
-M-
Measures 333
Movies 215